WHY GOOD LEADERS CHOOSE TO PLAY THE VILLAIN: THE EFFECTS OF MORAL LICENSING AND PERCEIVED TRUST ON LEADER BEHAVIOR By Zachary W. Woessner A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Psychology—Master of Arts 2022 ABSTRACT This study investigated why leaders who feel trusted by their followers might take license to perform immoral behaviors. Trust exists between agents who believe that the other party has their best interest in mind. This trust is established through social exchanges and has, in past studies, acted as a form of moral currency giving leaders license to behave poorly. This study found limited support for moral licensing when leaders perceive themselves to be trusted based on employee behaviors. Furthermore, I found little to no support for the effects of individual differences that might have informed how leaders would keep track of or choose to spend this moral currency through moderation effects via leaders’ motives and personality. This paper is to examine the process through which leaders develop a perception of being trusted by followers and if this perception mediates the relationship between employee and leader behaviors. Additional exploratory hypotheses suggest that the three factors of trust—ability, benevolence, and integrity—might matter when considering leader perceptions of feeling trusted by their followers, but the overall effects are relatively insignificant. Future studies should consider not only the effects of the three factors of trust, but also how trustors and trustees might respond to feeling trusted as the relationship continues to develop. To moments of grace and acts of perseverance. iii ACKNOWLEDGEMENTS Thank you. To Ben Woessner, for helping me build structure. To Valerie “Vernita” Newman, for enforcing that structure. To Katie Woessner and Kevin Pacelli, for mood checks and joy. To Julie Woessner, my mom, for unconditional love and support. To Linnea Vicari for solidarity and friendship. To Kaitlin Paxton Ward, for grounding me and helping me feel human. To Susannah Chandhok, for opening the door to what comes next. And to all the others along the way, thank you. iv TABLE OF CONTENTS INTRODUCTION………………………………………………………………………………...1 LITERATURE REVIEW…………………………………………………………………………7 METHODS………………………………………………………………………………………26 RESULTS………………………………………………………………………………………..36 DISCUSSION……………………………………………………………………………………47 CONCLUSION…………………………………………………………………………………..57 REFERENCES…………………………………………………………………………………108 APPENDIX……………………………………………………………………………………..120 v INTRODUCTION Trust in leadership has been studied for nearly 60 years (Ferrin & Dirks, 2002) and has become a key component in leadership theories. Trust is defined as a willingness to be vulnerable, and exists between dyads, individuals and teams, and individuals and organizations (Mayer, Davis, & Schoorman, 1995; Cropanzano, Anthony, Daniels, & Hall, 2017). Previous research has established trust as a critical component of strong workplace relationships (Scandura & Pellegrini, 2008; Colquitt, Scott, & LePine, 2007). Its development is related to a trustee’s perceived ability, benevolence, and integrity (Mayer et al., 1995). As such, trust is critical for social group functioning. It facilitates relationships in which followers and leaders believe one another will accomplish the tasks they set out to do, and in the workplace, is reinforced through successful interactions. This is known as dyadic trust and is fundamental to understanding the relationships formed by leader-follower pairs (Butler & Cantrel, 1984). Indeed, dyadic trust is often considered fundamental to reciprocal exchange behaviors between individuals (Butler & Cantrel, 1984; Dunbar, 2004). The development of trust through successful interactions can be understood through the framework of social exchange theory (SET), which has been studied since the 1960s (Colquitt et al., 2007; Cropanzano et al., 2017). SET is broad and examines how social relationships are built through repeated, successful, and often reciprocal interactions between individuals (Cropanzano & Mitchell, 2005; Dunbar, 2004). These interactions demonstrate an entity’s ability, benevolence, and integrity, often resulting in the formation of trust. The most critical of these three dimensions of trust within dyads for supporting social exchange is sometimes argued to be integrity (Butler & Cantrell, 1984). Integrity speaks to an individual’s adherence to a set of established norms or even morals. Therefore, SET is a useful tool for understanding how social 1 relationships progress from economic transactions to quality social exchanges within dyads (Cropanzano et al., 2017). This same logic can also help explain why these relationships are sustained over time. Leaders who receive positive outcomes from followers are likely motivated to reciprocate positive behaviors as encouragement for more of the same. Indeed, this effect has been studied before in leader-follower relationships and is known as the behavioral consistency effect (Johnson, Venus, Lanaj, Mao, & Chang, 2012; Zhu, May, & Avolio, 2004). Leader behaviors are described as consistent because of the self-reinforcing nature of the outcomes. In other words, when leaders receive positive outcomes from their good behavior, they are inclined to participate in more good behavior. This effect may be viewed as an extension of SET as once individuals within a dyadic relationship have built, developed, and are maintaining trust, they will have additional motivation to sustain positive behaviors because of the positive feedback these behaviors provide. Thus, behavioral consistency builds upon SET to potentially explain how individuals within a dyadic relationship build, develop, and maintain trust in order to continuously pursue positive outcomes from the sustained relationship (Butler & Cantrell, 1984). According to SET, trust is developed in both parties of the social exchange through reciprocal actions (Cropanzano et al., 2017). However, developing and reinforcing trust has been largely studied from the perspective of trustors. For example, research has predominantly focused on the factors contributing to how followers develop and maintain trust in leaders (Chen, Zhu, & Liu, 2019). This is problematic because trust is developed through relationships between individuals, meaning that both the perceptions of the followers as trustors and the leaders as trustees matter. This is the essence of dyadic relationships, regardless of the power dynamic between the individuals within it: the lived experiences of both parties matter (Dunbar, 2004). For this reason, theorists often frame trust development through the aforementioned SET (Chen 2 et al., 2019; Cropanzano et al., 2017) in which repeated social exchanges occur within dyads, or between trustors (followers) and trustees (leaders). In this context, an exchange is as simple as a leader requesting a follower to complete a task, and the follower granting that request. A shared history of successful exchanges establishes psychological safety for followers as more information about their leader’s ability, benevolence, and integrity becomes available (Mayer et al., 1995). Psychological safety is necessary for trust as it allows followers to be more comfortable with being vulnerable to their leader. In other words, leaders who have adequate ability, benevolence, and integrity are often considered trustworthy (Mayer et al., 1995). However, the role of integrity in trust development within dyads is potentially stronger in some contexts than others. This is noteworthy because integrity measures often address an individual’s morally consistent behavior (Mayer & Gavin, 2005; Butler & Cantrell, 1984). Therefore, in moral contexts, this dimension of trust within dyads might be more pronounced or account for more variance. Past research has broadly found mostly positive outcomes of follower trust in leaders such as lower intentions to quit and higher job satisfaction (Scandura & Pellegrini, 2008). Followers’ trust in leaders also contributes to better task performance and positive workplace behaviors like organizational citizenship behaviors (OCBs) (Scandura & Pellegrini, 2008; Gerstner & Day, 1997). Knowing this, it becomes increasingly clear that followers’ trust in leaders is important, if not critical, to workplace success. While understanding followers’ trust in leaders is important, this examines only one side of the relationship surrounding trust. When we view this relationship between leaders and followers from the leader’s perspective, we know much less. The antecedents and outcomes of leaders feeling trusted by followers have rarely been studied, if at all, from the leader’s perspective (Chen et al., 2019). This is a problem because leadership is constructed within 3 dyadic pairs—by both leaders and followers together (DeRue & Ashford, 2010). Therefore, not knowing how one party responds to trust, a critical component of the relationship, creates a serious gap in our understanding of subsequent leader behaviors such as abusive supervision, initiating structure, and consideration behavior (Tepper, Henle, Lambert, Giacalone, & Duffy, 2008; Podsakoff, Todor, & Skov, 1982; Korman, 1966; Bass & Stogdill, 1990). In this study, I will examine follower voice behavior and task performance as predictors of a leader’s perceived trust from followers due to their capacity to demonstrate a trustor’s belief in a trustee’s ability, benevolence, and integrity. Additionally, how the leader-follower relationship is perceived will vary between leaders. One explanation is that some leaders might keep track of social exchanges with followers more than others. This individual difference is defined as a leader’s reciprocity orientation or reciprocity norm (Cropanzano & Mitchell, 2005; Eisenberger, Lynch, Aselage, & Rohdieck, 2004). Inclusion of this individual difference is appropriate within the framework of the behavioral consistency effect because some leader behaviors have been found to be more consistent than others based on a leader’s personality (Funder & Colvin, 1991). For example, reciprocity orientation might change how a leader interprets their relationship, ultimately affecting their level of perceived trust from followers based on the predictors of employee voice behavior and task performance that later result in downstream behavioral differences for the leader (Cropanzano & Mitchell, 2005; Zhu et al., 2004). In summary, there is little knowledge about what predicts much of the variance in why a leader feels trusted by their followers, and even less information about how leaders’ perceptions of being trusted might be related to their behavior toward followers. Additionally, there is a difference in the literature between trusting and feeling trusted. Trust is conceptualized as a willingness to accept vulnerability to someone else. However, 4 feeling trusted is represented by an awareness of someone else’s willful vulnerability to oneself (Chen et al., 2019). While there has been some work on followers who feel trusted by their leaders (Brower, Lester, Korsgaard, & Dineen, 2009), there has been scarce work on leaders who feel trusted by their followers (Chen et al., 2019). Past studies have found that leaders who feel trusted by their followers will likely behave in one of two ways. First, trusted leaders might treat employees better, generating positive psychological experiences and accommodating employee interests and concerns (Cropanzano et al., 2017). This effect can be explained within the framework of social exchange theory and behavioral consistency (Zhu et al. 2004). Indeed, research shows that leaders who feel trusted by their followers engage in more subsequently positive behaviors such as OCBs (Podsakoff, MacKenzie, Moorman, & Fetter, 1990), consideration, and transformational behaviors (Dirks & Ferrin, 2002). These effects can be explained nicely by behavioral consistency because an individual continues to act in accordance with past behavior in order to produce similar future results; indeed, this stance has a long history of support in the literature (Johnson, Venus, Lanaj, Mao, & Chang, 2012; Funder & Colvin, 1991). However, existing research also reveals inconsistencies in how leaders act towards followers when positive relationships exist within leader-follower dyads. Some studies have found relationships suggesting that leaders’ who perceive themselves to be trusted might treat their followers worse and become lax in their contribution to leader-follower relationships because they feel secure in their position (Miller & Effron, 2010). For example, Lin, Ma, and Johnson (2016) found that leaders’ ego depletion and moral licensing had a positive relationship with the occurrence of unethical leader behaviors, and Ormiston & Wong (2013) found that positive leadership behaviors in high level leaders was associated with later, negative behaviors. This second effect contradicts SET and the behavioral consistency effect because leaders who 5 participate in good exchanges with followers would be expected to repay this trust with positive behaviors. Therefore, there is a gap in the literature as to why some leaders who feel trusted choose to behave poorly and others do not. To address this gap, I turn to moral licensing theory (Miller & Effron, 2010). In its most basic form, moral licensing theory states that good results eventually lead to poor outcomes (Miller & Effron, 2010). Essentially, performing a surplus of morally correct actions licenses an individual to perform subsequent immoral behaviors. This theory is important in the context of work because it offers an explanation for why otherwise high performing individuals might have behavioral lapses (Blanken, Van de Ven, & Zeelenber, 2015; Miller & Effron, 2010). Moral license can be taken in two distinct ways: through the spending of moral credits and the accruing of moral credentials. Credits are accumulated through morally good behavior and can be spent to license subsequently poorer ones (Miller & Effron, 2010; Lin, Ma, & Johnson, 2016). Credentials, or an established morally good reputation, may change the way that an individual views an action such that it is no longer unethical (Markus & Wurf, 1987). However, what researchers do not yet understand is how individual differences in motivation affect the frequency with which moral license is taken (Blanken et al., 2015; Miller & Effron, 2010). This is important because moral license is not always taken despite the accrual of moral credits and credentials (Merritt, Effron, & Monin, 2010). Given that past studies have found that individual differences in motivation are related to the behaviors individuals choose to perform, it is important to consider these same variables when attempting to understand how moral licensing’s effect may manifest more strongly in some individuals (Heckhausen & Heckhausen, 2008). Motivation orientations are likely one explanation because they determine what behaviors are important to an individual and vary between individuals (Boyatzis, 2016). According to 6 McClelland’s (1965) theory of motive acquisition, often referred to as McClelland’s theory of needs (Boyatzis, 2016), there are three primary motivation orientations or needs: the need for power (dominance motivation), the need for achievement (achievement motivation), and the need for affiliation (affiliative motivation). These motives might contribute to why leaders choose to engage in varying levels of behaviors of abusive supervision, initiating structure, and consideration behavior respectively. Past research on moral licensing has not considered how the relative strengths of these motives might impact the decision of individuals to take moral license through credits or credentials. This study aims to bridge this empirical gap in the literature on individual differences and moral licensing. Ultimately, this study strives to investigate the question of how a leader’s perceived trust from followers mediates the relationship between follower behaviors and leader behaviors. To address this, we must first understand how a leader develops this perceived trust from followers through two proposed predictors: employee voice behavior and employee task performance. The strength of these predictors’ effects on a leader’s perceived trust from followers will be moderated by a leader’s reciprocity orientation. A leader’s perceived trust from followers will then predict the following leadership behaviors: abusive supervision, initiating structure, and consideration behavior. The relationships between a leader’s perceived trust from followers and these three outcomes will be moderated by individual motivation orientations: the need for power, the need for achievement, and the need for affiliation respectively. LITERATURE REVIEW Throughout their history, trust and leadership have been studied in one of four ways: leader’s trust in followers, followers’ trust in a leader, leader’s perceived trust from followers, followers’ perceived trust from a leader (Chen et al., 2019). Of these four, trust is most typically 7 studied from the perspective of followers (Chen et al., 2019). This might be because leaders and employees exist within the hierarchy of organizations and low-power employees depend on high- power leaders for valuable information, good performance appraisals, and promotion (Chen et al., 2019). Perhaps it is because multiple followers are vulnerable to a single leader that research on how followers trust their leaders is anticipated to have most relevant relationship in the workplace. However, there are both formal and informal networks within an organization. As such, not all power comes from having a powerful title, but also having strong interpersonal networks (Brown & Eisenhardt, 1997). These multiple sources of potential power within organizations give employees the ability to undermine their leaders without necessarily using formal structures. Therefore, to limit the worry that leaders might feel regarding follower loyalty and compliance, it is important that they feel that followers both trust them and perceive them as a benevolent and capable leader. The lack of research on how leaders’ perceive trust from followers is a problem because trust exists within relationships—within leader-follower dyads. As such, followers’ trust in leaders impacts leader behavior, but we do not yet understand the extent of this impact. This is important for understanding the theoretical underpinnings of trust in leader-follower relationships, but also optimal leader behaviors to build trust and engender positive outcomes in the workplace (Ferrin & Dirks, 2002; Gerstner & Day, 1997). As previously mentioned, trust is defined as the willingness of a trustor to be vulnerable to a trustee based on the trustee’s ability, benevolence, and integrity (Mayer et al., 1995). Benevolence is an understanding that the trustee has one’s best interest at heart. Integrity refers to a trustee’s “adherence to a set of principles that the trustor accepts” (Mayer et al., 1998, p. 719). And lastly, ability refers to whether a trustee is perceived to be competent. However, this proposed research studies trust from the perspective of the trustee. Studying trust from the 8 perspective of the trustee is necessary because trustees act on the trust they receive. Therefore, to fully understand the relationship between trustees and trustors, we must understand the behavioral processes following the development of trust. However, this study frames trust through the lens of moral licensing. As such, this study is particularly concerned with the integrity dimension of trust as it might account for morally consistent behavior compared to ability and benevolence. To understand how trust develops, we must first understand the antecedents of a trustee feeling trusted by a trustor. In the proposed study, I will examine leaders’ perceptions of being trusted by their followers. Therefore, the leader is the trustee and the follower is the trustor. Because trust research has largely focused on followers in the leader-follower dyad, there is scarce information regarding what might predict or be related to a leader’s perceived trust from employees. To address this gap in the literature surrounding the predictors of leaders feeling trusted by followers, this study proposes to view the effects of some common outcomes of followers’ trust in leaders, such as employee voice behavior (Botero & Van Dyne, 2009) and employee task performance (Colquitt et al., 2007), from the perspective of leaders. In other words, because followers’ trust in a leader predicts higher levels of employee voice behavior and higher employee task performance (Scandura & Pellegrini, 2008), I argue that leaders might use their perceptions of an employee’s voice behavior and task performance as predictors for how trusted they are by their followers. In the following paragraphs, I explain why these two behaviors are indicative of an employee trusting their leader as it relates to a leader’s ability, benevolence, and integrity. Employee voice behavior and task performance can be framed with social exchange theory (SET) as a means for explaining the proposed leader-follower relationship through trust. 9 SET is a broad lens for examining how social relationships are built through repeated, successful interactions between individuals (Cropanzano & Mitchell, 2005; Cropanzano, Anthony, Daniels, & Hall, 2017). These interactions are social exchanges in which a request is made from one individual to another; successful completion of requests builds the social relationship. The literature often cites higher levels of trust and stronger relationships as a result of social exchanges (Cropanzano & Mitchell, 2005). As such, it is therefore logical to argue that through repeated, successful interactions individuals demonstrate ability, benevolence, and integrity in their workplace relationships. Social exchange theory can therefore serve as a theoretical framework through which we can examine the development of trust between leaders and followers. We can interpret the outcomes of employees trusting a leader (e.g., voice behavior and task performance) as potential antecedents of a leader’s perceived sense of trust from followers when viewed through the framework of social exchange theory (SET). Employee Voice Behavior The first of these two antecedents is employee voice behavior. Voice behavior is defined as an employee’s willingness to express a desire for change or discomfort with the current state (Morrison, 2014; Detert, Burris, Harrison, & Martin, 2013; Van Dyne et al., 2008). While voice can be promotive (an improvement) or prohibitive (a complaint) (Botero & Van Dyne, 2009), it ultimately constitutes a risk for followers because it often asks the leader to engage in change- related actions. This request is a key component of voice behavior as it challenges the leader’s power for a productive reason. Therefore, voice behavior requires a degree of vulnerability when expressed from followers to leaders since the leader can either support the idea or punish the follower for their dissenting opinion. As such, followers must believe that their leaders both have the ability to fix the current situation, but that they are also benevolent and will not retaliate 10 against the employee for speaking up. As the trustee, leaders will look for signs that they are perceived as competent, benevolent, and have integrity in the eyes of their trustor. The dimension of integrity is likely especially important for framing leader’s behaviors as consistent. Followers’ voice behavior addresses these facets of trust, thereby displaying follower trust in the leader (Mayer et al., 1995). Employee Task Performance The second proposed antecedent of a leader’s perceived trust from followers is an employee’s task performance. Task performance is defined as how well an individual performs their assigned tasks at work (Anand, Liden, Vidyarthi, 2011; Colquitt et al., 2007; William & Anderson, 1991). Employee task performance demonstrates a follower’s trust in a leader in two ways. Because a leader often presents the work of their followers to stakeholders and higherups, how much they label an employee’s work as their own is likely a function of a leader’s integrity and benevolence. If the leader shares successes fairly and an employee believes that the leader will applaud them in front of stakeholders when deserved, they are likely to work harder as they will see more of the potential rewards. Leaders who perceive a follower to be performing their tasks successfully might then interpret this to mean that their follower perceives that the leader has integrity such that they adhere to a moral code, and benevolence such that they have the follower’s best interest at heart. Secondly, if a leader lacked the ability to do their job, a follower might not be motivated to ‘pick up the slack’ or work hard themselves. Past research supports this argument and has found that a follower who does not believe in their leader’s ability or trust their leader’s directions has been found to withhold task performance (Colquitt et al., 2007). Because of this, a follower who perceives their leaders as competent are more likely to perform tasks well. Therefore, leaders likely perceive a follower who performs their tasks to the best of 11 their ability as trusting of the leader. Because both voice behavior and task performance demonstrate a follower’s potential belief in a leader’s ability, benevolence, and integrity, these two follower behaviors are hypothesized as predictors of a leader’s perceived trust from the follower. H1: Follower’s task performance will be positively related to leader’s perceived trust from a follower. H2: Follower’s voice behavior will be positively related to leader’s perceived trust from a follower. However, people vary in how they make attributions about the world based on their own values and motives (Parker, 1998). As such, I propose that the relationship between employee voice behavior and employee task performance with a leader’s perceived trust from a follower will be moderated by the leader’s reciprocity orientation. Reciprocity orientation is defined as the extent to which an individual tracks or participates in “score keeping” of obligations (Cropanzano & Mitchell, 2005; Molm, 2003; Yoshikawa, Wu, & Lee, 2019). Past studies have found that not all individuals value reciprocity in the same way (Rousseau & Schalk, 2000; Shore & Coyle-Shapiro, 2003). Specifically, leaders who are high on reciprocity orientation will value adherence to reciprocal exchange in their relationships (Ackerman, Fleiẞ, & Murphy, 2016). This is in line with the behavioral consistency effect as well (Funder & Colvin, 1991). Because of this, they will pay more attention to what they request and what they are granted to ensure that the exchanges are reciprocal or equal. I propose that this heightened attention to equal reciprocal exchanges will incline leaders with high reciprocity orientations to take more notice of and care more about follower behaviors such as task performance and voice. This increased 12 attention will therefore strengthen the relationship between the proposed predictors and a leader’s perceived trust from a follower. As such, a leader’s reciprocity orientation will moderate the relationship between the proposed antecedents of trust (employee voice & task performance) such that the positive relationship between the antecedents and a leader’s perceived trust will be stronger when reciprocity orientation is high and weaker when it is low. This means that those high in reciprocity orientation will be more sensitive to the give and take in their relationships, ultimately affecting the levels of trust leaders perceive from a follower. H3: Leader’s reciprocity orientation will moderate the positive relationship between a follower’s (a) task performance and (b) voice behavior with the leader’s perceived trust from the follower. Specifically, the positive relationships between a follower’s task performance and voice behavior with leader’s perceived trust from the follower will be stronger when the leader’s reciprocity orientation is high versus low. Outcomes of Leaders’ Perceived Trust Leaders who perceive trust from a follower may be more likely to engage in a variety of different leader behaviors. In the current research, I will explore how leaders’ perceived trust relates to leader behaviors including abusive supervision, contingent reward, and consideration. Broadly understood, trust is established through repeated interactions between individuals (Cropanzano & Mitchell, 2005). These quality relationships are developed through these repeated interdependent exchanges between leaders and followers in an effort to yield positive individual outcomes for both parties of the dyad (Gerstner & Day, 1995). Thusly reinforcing my earlier argument that trust matters from the perspective of both involved parties. Therefore, to understand how trust impacts the relationship between leaders and followers, we must examine it 13 from the perspective of leaders as well. While trust is normally a good thing, I argue that it might not always lead to the best outcomes. However, the theory of moral licensing can explain why good behaviors become bad. In its simplest form, moral licensing is defined as when good behaviors result in subsequently bad consequences (Miller & Effron, 2010). Moral licensing theory can be broken down into two sub-components: moral credits and moral credentials. These two components explain the process through which good behavior results in bad consequences. Moral credits are a resource accrued by performing morally laudable behavior (Miller & Effron, 2010). As previously described, these credits function as a currency through which moral licensing can be transacted. In trading in a credit built up from earlier good behavior, an individual might be able to perform an instance of bad behavior with little or no lasting repercussions. The second way moral licensing can occur is through moral credentials. Moral credentials are a measure of one’s moral self-regard (Markus & Wurf, 1987) and are related to an individual’s character. In other words, they influence our interpretations and attributions about an event such that it is not viewed as immoral. A leader feeling that they are a good person and relying on this reputation to dismiss bad behavior is an example of moral credentials. In such a case, credentials are established through the development of an overall positive reputation, and negative behaviors that would otherwise produce dissonance with this reputation are dismissed. Moral credentialing is not an explicit process. Rather, it fundamentally changes how individuals view the morality of their actions based on their past behavior. Moral licensing occurs through these two pathways such that once an individual has accrued a surplus of moral currency, they may be licensed to refrain from good behavior (Lin et al., 2016; Miller & Effron, 2010; Sachdeva, Iliev, & Medin, 2009). 14 The question I must next address is how do leaders build these moral credits and moral credentials? I argue that leaders accrue moral currency through the previously highlighted SET. This shared history of social exchanges builds trust (Cropanzano & Mitchell, 2005). Specifically, leaders likely rely on the integrity dimension of trust when taking moral license. Because of this shared, generally moral history, there is less risk in being vulnerable to another. Therefore, moral licensing argues that once this relationship has been built through social exchange, this sense of trust might give leaders license to behave poorly because the consequences of negative behaviors will be lessened by the presence of surplus moral credits and credentials, ultimately not risking the relationship as a whole despite behaving poorly. However, how leaders choose to take moral license will vary from individual to individual. This study will examine the effect of individual differences on moral licensing through McClelland’s theory of motive acquisition. According to McClelland (1965), there are three basic needs that motivate much of human behavior: power, achievement, and affiliation. The need for power is an unconscious drive to have an impact on others (Boyatzis, 2016); the need for achievement is an unconscious drive toward a standard of excellence (Boyatzis, 2016); and the need for affiliation is an unconscious drive to have warm, close relationships (Boyatzis, 2016). These three motives have been found to vary in strength across individuals (McClelland, 1965; Boyatzis, 2016), and therefore contribute to behavioral differences between leaders. I argue that these motives will be related to when a leader chooses to take moral license or not. The three leader behaviors in the proposed model are matched with motives in McClelland’s (1965) theory of needs, as expanded upon below. The stronger a need, the more or less likely a behavior is to occur. Abusive supervision is likely motivated by an individual’s need for power or power motivation (Tepper, 2000). Initiating structure is likely motivated by an individual’s 15 need for achievement (Nicholls, 1984; Judge, Piccolo, & Ilies, 2004). Consideration behavior is likely motivated by an individual’s need for affiliation. I argue that moral license will be taken differently by different leaders based on these three motivation orientations and that leaders’ who feel trusted will engage in more abusive supervision, less initiating structure, and fewer consideration behaviors. I expand upon each of these behaviors and its subsequent moderator in the following paragraphs. Moreover, I argue that the frequency at which leaders take moral license will vary between individual leaders because of differences in these motives (McClelland, 1965; Boyatzis, 2016; Miner, 2005). That is, these three leader motivations will each moderate the strength of the relationship between a leader’s perceived trust from followers and one of three leader behaviors: abusive supervision, initiating structure, and consideration. This contribution is meaningful because the literature around moral licensing does not usually consider individual differences (Miller & Effron, 2010). We ought to pay more attention to how people think about their moral credits and credentials to better understand this process. I argue that how leaders value and view their relationships with their followers will be related to how likely they are to take moral license. In the following paragraphs, I hypothesize the main effect of a leader’s perceived trust from followers on each of the three aforementioned behavioral outcomes (abusive supervision, initiating structure, and affiliative behaviors) as would be predicted under moral licensing. I also hypothesize that the relationship between a leader’s perceived trust from followers with each of these outcomes is moderated by a motive from McClelland’s theory of needs (1965) and explain how this moderation will be related to when and why leaders choose to take moral license. The first outcome that is predicted by a leader’s perceived trust from followers is abusive supervision. Abusive supervision is defined by Tepper (2000) as the sustained display of hostile 16 verbal and nonverbal behaviors. In the context of trust, abusive leadership behaviors are most frequently interpersonal in nature. Moral licensing can explain why these negative leader behaviors are performed despite the positive shared history between a leader and follower. When leaders perceive themselves as trusted, this might act as a form of moral credit, allowing leaders to spend the credit on less moral behaviors. Alternatively, feeling trusted might provide leaders with the moral credentials needed to dismiss the immorality of a behavior outright. As such, moral licensing can potentially explain why leaders might engage in abusive supervision. Indeed, past studies have found an association between abusive supervision and moral licensing (Lin et al., 2016). Because of this, I argue that the main effect of a leader’s perceived trust from followers on abusive supervision will be positive. H4: Leader’s perceived trust from a follower will be positively related to her/his abusive supervision towards the follower. However, the extent to which a leader’s perceived trust from followers will be related to abusive supervision will be moderated by a leader’s need for power. An individual’s need for power is a motivation orientation and individual difference. It is defined as motivation to feel powerful, strong, or dominant when engaged in an activity (Busch, 2018) or simply put, having the desire to influence others (McClelland, Koestner, & Weinberger, 1989). It is sometimes referred to as dominance motivation or the need for power interchangeably in the literature (Busch, 2018; Heckhausen, 1967). As such, it plays off the power dynamic that exists between individuals (Busch, 2018). This matters in the context of trustors and trustees as leaders have more power as the trustee (Dirks & Ferrin, 2002). I argue that leaders who have a high need for power compared to low will have a stronger positive relationship between a leader’s perceived trust from followers and abusive supervision. This is because leaders who have a high need for 17 power will choose to take moral license more often than leaders who have a low need for power. Leaders with a high need for power will view their moral credits as something to be spent to feel powerful, therefore higher levels will strengthen the positive relationship between leader’s perceived trust from followers and abusive supervision. This logic is supported by past studies that found a positive relationship between an individual’s need for power and their acceptance of abusive behaviors as normal or acceptable (Martinko, Harvey, Brees, & Mackey, 2013; Tepper, Simon, & Park, 2017). Alternatively, a leader will have also built moral credentials as someone who is benevolent, competent, and has integrity in the workplace. Integrity, as it relates to moral consistency, might be especially important to consider here. These credentials allow leaders to reframe unethical behaviors as ethical or in line with their own self-image as an upstanding, moral person (e.g., someone with integrity), especially if high follower performance is maintained. Therefore, I propose that leaders who have a high need for power will see a stronger relationship between perceived trust from followers and abusive supervision behaviors compared to those who are low. H5: A leader’s need for power will moderate the relationship between leader’s perceived trust from a follower with abusive supervision such that the positive relationship between the two will be stronger when the need for power is high versus low. Not only will leaders perform more negative behaviors, but I argue that moral credits and credentials will give leaders license to perform fewer positive behaviors. An example of this is initiating structure, a transactional leadership behavior commonly used to motivate followers. Initiating structure reflects “the extent to which an individual is likely to define and structure his role and those of his subordinates toward goal attainment” (Korman, 1966, p. 349). Indeed, a meta-analysis by Judge, Piccolo, and Ilies (2004) found that initiating structure is related to 18 follower job satisfaction, follower satisfaction with the leader, follower motivation, and leader effectiveness. In other words, initiating structure is a leadership strategy used to strengthen positive employee workplace outcomes. Moral licensing can explain why leaders who feel trusted by their followers might perform fewer initiating structure behaviors. Much like the argument for moral licensing’s effect on abusive supervision, a leader’s perceived trust from followers might give them license to perform fewer initiating structure behaviors for similar reasons. Feeling trusted might act as a form of moral credit through accrued trust or change how leaders view themselves, ultimately changing how they interpret the morality of their behaviors (e.g., because I’m a good, trusted leader, my behaviors are good too). Perhaps trust as a form of moral credits can be spent to deviate from good leadership behaviors. Because initiating structure is a generally effective tool for leaders to motivate employees (Judge et al., 2004), moral licensing explains that leaders who feel trusted might not see the need to engage in more initiating structure behaviors. Alternatively, moral credentials reframe unethical actions such that they are not even construed as unethical. These credentials might then allow leaders to withdraw positive behavior like initiating structure. H6: Leader’s perceived trust from a follower will be negatively related to her/his initiating structure behavior towards the follower. The extent to which a leader’s perceived trust from followers will be related to initiating structure will be moderated by a leader’s need for achievement. The need for achievement is defined by Nicholls (1984) as behavior striving to demonstrate high rather than low ability. Others view the need for achievement as the desire to instill the perception of competence (Nicholls, 1984; Heckhausen, 1967). Podsakoff et al. (1982) found that high performing employees are more motivated by transactional leadership behaviors. Therefore, they will 19 produce more work when rewarded adequately, thus maintaining a high level of function. Because individuals are motivated in different ways, it would make sense that leaders who value high achievement would want each follower to continue performing at their peak level, subsequently increasing the leader’s own effectiveness. I argue that a high need for achievement will weaken the negative relationship between a leader’s perceived trust from a follower and their initiating structure behaviors compared to when it is low. I argue that moral credits and credentials will be used to perform initiating structure behaviors differently by people with high versus low need for achievement. An individual’s need for achievement will change the way that leaders choose to spend their moral credits. Leaders who have a high need for achievement will view initiating structure as an important leader behavior (Van Eden, Cilliers, & Van Deventer, 2008), perhaps too important to be forgone through moral licensing. In other words, because leaders with a high need for achievement value high performance in their own work and their team’s, anything that detracts from either is too expensive a pursuit. Those with low need for achievement will see less of this effect. The need for achievement will also change the way that leaders feel about their moral credentials. Those with high need for achievement will view themselves as good leaders because they engage in behaviors that motivate their followers to perform well (Margerison, 1984), thereby satisfying their own need for achievement. Choosing not to engage in initiating structure might then reflect poorly on their character. Therefore, when leaders with high need for achievement have accrued moral credits and credentials, they will be less likely to take moral license because they value a consistent, high level of performance (Podsakoff et al., 1982; Miller & Effron, 2010) and utilizing such a resource would detract from their achievement. Meanwhile, leaders with low need for achievement will be more inclined to 20 spend their moral credits and credentials because they do not value achievement in the same way. H7: Leader’s need for achievement will moderate the relationship between leader’s perceived trust from a follower with her/his initiating structure such that the negative relationship between the two will be weaker when need for achievement is high versus low. The final leader behavior that is influenced by a leader’s perceived trust from followers is consideration. Consideration is defined as the “extent to which an individual has relationships characterized by mutual trust, respect for subordinates’ ideas, and consideration of their feelings” (Korman, 1966, p. 349; Kerr & Schriesheim, 1974). Indeed, past research has found that leader consideration behaviors show concern and respect for followers (Bass & Stogdill, 1990) and are generally positive, focusing on follower development (Bass & Avolio, 1997). I argue that leaders who perceive themselves to be trusted by their followers will engage in less consideration behavior. As with the previous outcome, I again argue that this effect can be explained by moral licensing theory (Miller & Effron, 2010). Moral licensing occurs through a buildup of moral credits and credentials. A leader’s perceived trust from followers, as previously argued, might act as moral credit. This credit might then be spent to engage in fewer consideration behaviors—an arguably poor outcome of feeling trusted. However, moral licensing describes when good things go bad (Miller & Effron, 2010), thusly supporting this argument. Similarly, leaders might perceive themselves to be good leaders, having strong moral credentials because they feel trusted. This feeling of being a good, trusted leader might give leaders the impression that they can put less effort into follower relationships because they already feel trusted, thereby reducing their consideration behavior. 21 H8: Leader’s perceived trust from a follower will be negatively related to his/her consideration behaviors towards the follower. However, some people value social relationships more than others. This is known the need for affiliation. Need for affiliation is an individual motivation orientation dealing with how people are motivated to engage in social relationships (Heckhausen, 1967). Oftentimes, this means that people are motivated by “feeling warm and friendly,” or having “social interaction during an activity” (Heckhausen, 1967). Indeed, past research has found that the need for affiliation generally fosters positive relationships between individuals as it promotes benevolence and integrity in interpersonal interactions, two facets critical to trust (Mayer et al., 1995; Heckhausen, 1967). In the context of consideration behaviors, the need for affiliation will likely moderate the negative relationship between a leader’s perceived trust from followers on consideration such that it will be weaker when the need for affiliation is high compared to when it is low. Moral licensing can explain this effect. Similarly to the two previous moderation arguments for dominance and achievement motivations, the need for affiliation will change how moral credits are spent and how moral credentials are perceived by leaders. As previously argued, leaders accrue and possess moral credits in the form of perceived trust. Since the need for affiliation describes how strongly an individual is motivated by interpersonal relationships, individuals with a high need for affiliation will value relationships with followers more than those with low. The need for affiliation matters in the context of moral licensing such that leaders who have a high need for affiliation will likely value their status as a benevolent leader who has integrity in their workplace relationships to such an extent that they would not consider being less considerate toward their followers—even if they could. Therefore, the appeal of spending moral credits through moral licensing, even if accrued, is lessened by this motivation 22 orientation. At the same time, leaders possess moral credentials from feeling trusted by followers (e.g., because they trust me, I must be a good person). Because these same people who have a high need for affiliation value their status as benevolent leaders, they will also be less likely to use moral credentials to reduce their consideration behavior. While this status could be used to perform fewer consideration behaviors, leaders with a high need for affiliation will be unable to reframe performing fewer good leadership behaviors as ethical because of how highly they value their relationships with employees. To do so would likely cause stress in the form of cognitive dissonance. In summary, moral licensing theory predicts that because the leader is otherwise good, lapses in good behavior might be tolerated. However, leaders with a high need for affiliation will see a weaker relationship between a leader’s perceived trust from followers and their consideration behavior compared to those who are low. As such, they will be less likely to take moral license to perform fewer consideration behaviors because they value relationships too highly. H9: Leader’s need for affiliation will moderate the relationship between leader’s perceived trust from a follower with consideration such that the negative relationship between the two will be weaker when the need for affiliation is high versus low. Additionally, I propose two exploratory hypotheses regarding leaders’ workplace incivilities as an alternative outcome for a leader’s abusive supervision behavior. According to Tepper et al. (2017), individuals who engage in abusive supervision often do not often consider their behaviors to be abuse. As such, a milder form of employee mistreatment might better capture the effects of moral licensing under this self-report study design. 23 H10a: Leader’s perceived trust from a follower will be positively related to incivility towards the follower. H10b: A leader’s need for power will moderate the relationship between leader’s perceived trust from a follower with incivility such that the positive relationship between the two will be stronger the need for power is high versus low. While I expect to see positive small to medium effect sizes in support of my hypotheses, it could be the case that leaders do not take moral license when they perceive themselves to be trusted. SET would suggest that leaders would not change their behavior for fear of damaging their relationship and negatively affecting future social exchanges. This perspective is further supported by the behavioral consistency effect (Funder & Colvin, 1991; Johnson et al., 2012; Zhu et al. 2004). Instead, leaders might build idiosyncratic credits from perceived employee trust as opposed to moral credits or credentials. Idiosyncratic credit is broader than moral license such that it encompasses any status or membership in a group, and when this credit is depleted, an individual is no longer a member of said group (Hollander, 1968). As such, one ought to be able to build idiosyncratic credit through social exchanges in the same way that one might build moral credit within dyads of leaders and followers. Furthermore, idiosyncratic credit does not need to be spent solely on negative behaviors as is the case of moral licensing. Instead, leaders can spend idiosyncratic credits to encourage followers to engage in positive behaviors. Because idiosyncratic credits represent status in a group, they can be operationalized as an individual’s leader identity, or to what extent an individual is one of the group, works for the group, crafts an identity for the group, and makes that identity matter to the group (Steffens, Haslam, Reicher, Platow, Fransen, Yang, Ryan, Jetten, Peters, & Boen, 2014). Leaders who identify strongly with these four dimensions of leadership identity are likely to have more influence over their 24 followers’ actions (e.g., more idiosyncratic credits) and potentially more consistent leadership behaviors (Johnson et al., 2012). In the context of this study, as leaders feel more trusted (a good perception), they will likely feel obligated to repay that good felling with more good behaviors and fewer bad ones. However, leaders do not always behave in a consistently good way. Ormiston and Wong (2013) found positive behaviors in high-level leaders are sometimes associated with subsequent negative behaviors. A strong leader identity and the behavioral consistency effect would suggest that the opposite should occur (Funder & Colvin, 1991). Instead, Ormiston and Wong’s (2013) contradictory findings might potentially be explained by an individual’s moral identity. Moral identity is described by Aquino and Reed (2002) as part of a person’s self-concept that informs their cognitions and actions based on internalized social identities of morality. In other words, an individual’s moral identity can be used to decide the difference between what one should do and what one must do to remain a moral person (Aquino & Reed, 2002). For example, some leaders feel they must give employees bonuses around the holidays if the company had a good year. Other leaders might feel that they should, but that it is their decision if they do. As such, leaders with weaker moral identities will have a wider range of behaviors that they can choose to do because there are fewer behaviors that they absolutely must perform. Thus, a leader’s moral identity likely informs what behaviors they will allow themselves to get away with and what behaviors they would never consider. Moral licensing supports this argument such that leaders who feel that there is less they should do, and potentially more they can do, will be more likely to take moral license to reduce positive behaviors (fewer must behaviors) and increase negative behaviors (more optional should behaviors). 25 Taken together, these two theories attempt to address the same phenomenon: how leaders behave when they feel trusted in follower relationships. SET and the behavioral consistency effect argue that the more a leader identifies as a leader, the more behaviorally positive and consistent behavior they will perform. Meanwhile, moral licensing theory argues that the weaker a leader’s moral identity, the more likely they are to take license and perform subsequently negative behaviors despite feeling trusted by their followers. This study will explore both theories as potential explanations for how leaders behave when they perceive trust from a follower. Furthermore, I will include additional outcomes (e.g., ethical leadership, Brown, Trevino, & Harrison, 2004; Laissez-faire leadership, Bass & Avolio, 2004) to broaden the possible leader behaviors I measure that might be informed by either theoretical framework. The data will be analyzed using the same methods as outlined below (e.g., SEM tested for model fit) and compared to the proposed moral licensing results. METHODS Participants Participants were recruited through Prolific.co using their custom screening function, which allows recruitment based on participant demographics. The population was limited to individuals who had completed over 150 previous requests at or above a 90% success rate (Keith, Tay, & Harms, 2017). I needed at least 228 participants to achieve power at .80 over the course of 3 surveys administered across 15 days (Preacher & Coffman, 2006. Two-hundred and ninety-nine participants were recruited for wave 1 and 258 completed all three surveys. There was a 9.00% attrition rate after the first wave and a 5.14% attrition rate after the second wave. Participants completed the study between November 6 and November 29 of 2020. Participants were reminded via Prolific’s internal messaging system to complete waves 2 and 3 of the survey 26 twice each week, the first was sent at midnight on the day the survey opened (Monday), the second was sent at the end of the workday on Monday. Participants could respond for 32 hours after the survey opened at midnight on Monday (12:00am GMT). Of the 258 participants who completed all three waves, 6 were excluded. Four were excluded due to acquiescent responses (e.g., answering the same value such as “Neither agree nor disagree” or “5” on a 10-point scale for all questions) and two for improbably short response times (e.g., those who completed the survey in less than 3 minutes – faster than the survey generator could pilot it without reading questions). Participants were compensated $2.37 per survey with an additional bonus of $3 for completing all three waves. Completing all three waves yielded $10.12 per participant. Proflific.co charges a 33% commission rate that is not included in these numbers. All participants worked full time (35 hours per week), were in a leadership role or had supervisory duties, had at least 2-3 direct reports (31% had 2-3, 28.2% had 4-6, 17.1% had 7-10, 23.8% had more than 10), interacted with their direct reports at work regularly, and lived in Europe or Canada. Prolific.co allows participants to return a request should they not meet the screening criteria. As such, most participants who did not meet the aforementioned demographics were pre-screened from the study. Of the remaining 252 participants, 56.7% identified as male (42.9% female, 0.4% other), 86.6% were white, 4.4% South Asian, 2.8% East Asian, 2.8% Black / African American, 1.6% Middle Eastern / Arab, and 2% other. Participants ranged in age from 23 to 68 with the average age being 40 (SD = 10.08). The average job tenure was 9.45 years (SD = 7.42) with 48.6% of employees working remotely 5 days per week and 97.1% working remotely 4 or more days per week during the global COVID-19 pandemic. The average organization size was 10,744 employees (median = 400 employees). The average 27 number of interactions leaders had with their direct reports was 76.7 per week (median = 40 interactions per week). Procedures Three surveys were administered to participants over the span of 15 days including two weekends. Once the survey was opened, participants had 1.5 hours to complete it with a seven- day lag between each administration. A limited response time is required through Prolific’s survey administration platform. Surveys opened at midnight (12:00am GMT) on Monday and closed 32 hours later, or roughly the start of the next business day GMT (see Table 1). A complete list of variables can be found in the following measures section. The predictors (employee voice and employee task performance) and individual differences (needs for power, achievement, affiliation; moral identity; and leadership identity) were measured at time one (T1). The leader’s perceived trust from followers was measured at time two (T2). And the outcome variables of leader behaviors (consideration, initiating structure, abusive supervision, incivilities, laissez-faire, and ethical leadership) were measured at time three (T3). These variables and their hypothesized relationships are seen in Figure 1. Additionally, each survey (Surveys 1, 2, and 3) asked the leader to name 3 followers. One of those three names was randomly selected and inserted into the behavior-based measures. Details on these adaptations can be found in the measures section. Measures were presented in blocks as their respective scales, and blocks were presented randomly to improve the validity of findings and potentially reduce socially desirable responses (Lensvelt-Mulders et al., 2005). Participants were able to select the request from Prolific.co should they have met previously described screening conditions. Only participants who successfully completed 28 previous waves were eligible to select the next request for the subsequent wave of the study. Participants were reminded of the longitudinal nature of the study in the consent documents, the survey conclusion, and two weekly email reminders. Both emails thanked participants for their time, reminded them of the survey to complete, and restated the monetary incentive to complete all three waves. As previously noted, the first email invited successful participants to the next wave (sent at midnight) and the second reminded them to complete the survey before the beginning of the next workday (sent end of day). Participants contacted the researcher through Prolific.co’s internal messaging feature to resolve technical difficulties with the survey (e.g., no record of completion on Prolific.co while full completion record existed in Qualtrics). There were fewer than 10 instances in which the researcher had to provide technical assistance to participants. All communication occurred through Prolific.co’s internal messaging system (recruitment, invitations, technical support, and payment confirmations). Measures Demographics. Demographic information was collected in Time 1. Participants reported their gender, sex, age, sexual orientation, ethnicity, education, fulltime status, job tenure, leadership role, size of the organization, job title, number of direct reports, and frequency of communication with direct reports. Employee Voice Behavior. I adapted LePine and Van Dyne’s (1998) scale to measure employee voice behavior from the leader’s perspective. The scale has had 6 items (α = .90; LePine & Van Dyne, 1998). An example adapted item is “[Employee] spoke up and encouraged others in the unit to get involved in issues that affected our work.” Leaders reported how frequently the employee performed each behavior using a slider from 0 to 10+ times in the last week. 29 Employee Task Performance. I adapted items from William and Anderson’s (1991) scale for in-role behaviors as a measure of task performance. There are 7 items (α = .81). An example item “[Employee] met formal performance requirements of the job.” Leaders reported how frequently the follower performed each behavior using a slider from 0 to 10+ times in the last week. Leader Reciprocity Orientation. I adapted Eisenberger et al.’s (2004) scale to measure positive reciprocity orientation. The scale assesses how individuals value retribution or reciprocation in social relationships (Eisenberger et al., 2004). The positive scale has 10 items (α = .85; Eisenberger et al., 2004) similar to “I feel uncomfortable when someone does me a favor that I won’t be able to return.” Leaders the extent to which they agreed or disagreed with each statement (from 1 = Strongly disagree to 5 = Strongly agree). Leader’s Perceived Trust From Followers. I adapted Mayer and Davis’s (1999) scale to measure perceived trust using its three dimensions: ability (6 items: α = .93), benevolence (5 items: α = .90), and integrity (6 items: α = .66). The overall reliability for all three scales measuring trust as one construct is α = .92. An example item from each scale: “[Employee] believes I am very capable of performing my job” (ability); “[Employee] believes I am very concerned about their welfare” (benevolence); “[Employee] never has to wonder whether I will stick to my word” (integrity). Leaders reported how they felt one of their followers would respond to each statement about their expectations for their leader’s behavior (from 1 = never to 5 = always). Abusive Supervision. I adopted Tepper’s (2000; Tepper et al, 2008) scale to measure abusive supervision from the leader’s perspective instead of followers. The scale has 15 items (α 30 = .77). An example item is “I put [Employee] down in front of others.” Leaders reported how frequently they performed each behavior using a slider from 0 to 10+ times in the last week. Initiating Structure. I adopted the LBDQ-Form XII scale to measure initiating structure (Ohio State Leadership Studies, 1962). It has 15 items (α = .92, Ohio State Leadership Studies, 1962). An example item is “I ruled with an iron hand.” Leaders reported how frequently they performed each behavior using a slider from 0 to 10+ times in the last week. Laissez-fair Leadership. I adopted the scale used by Bass and Avolio (2004) to measures laissez-faire leadership. It has 4 items (α = 76; Bass & Avolio, 2004). An example item is “I avoided getting involved when important issues arose.” Leaders reported how frequently they performed each behavior using a slider from 0 to 10+ times in the last week. Ethical Leadership. I adapted the scale used by Brown, Trevino, and Harrison (2004) to measure ethical leadership. It has 10 items (α = .76; Brown et al., 2004). An example item is “I discussed business ethics or values with [Employee].” Leaders reported how frequently they performed each behavior using a slider from 0 to 10+ times in the last week. Moral Identity. I adopted the scale used by Aquino and Reed (2002) to measure moral identity. It has 12 items (α = .89). Leaders responded to items based on a list of provided adjectives. An example item is “Having these characteristics is an important part of my sense of self.” Leaders rated how much they agree or disagree with each statement about themselves (1 = strongly disagree to 5 = strongly agree). Leader Identity. I adopted the scale by Steffens et al. (2014) to measure leader identity. It has 15 items with four sub scales (identity prototypicality, α = .91; identity advancement, α = .87; identity entrepreneurship, α = .89, identity impresarioship or embeddedness, α = .83). The 31 reliability for the full scale was α = .94. Example items for each scale respectively are “I am representative of members of my group,” “I promote the interests of members of the group,” “I develop an understanding of what it means to be a member of the group,” “I devise activities that bring the group together.” Leaders rated the extent to which they believe the scales describe them (1 = not at all to 5 = completely). Consideration. I adopted the LBDQ-Form XII scale to measure consideration behavior (Bass & Stogdill, 1990). It has 15 items (α = .96, Ohio State Leadership Studies, 1962). An example item is “I looked out for the personal welfare of [Employee].” Leaders report how frequently they performed each behavior within the past week using a slider ranging from 0 to 10+. Need for power. I adopted the IPIP CAT-PD scale (Simms et al., 2011) for need for power, also known as dominance motivation. It has 6 items (α = .84). An example item is “Have a strong need for power.” Leaders rated how much they agree or disagree with each statement about themselves (1 = strongly disagree to 5 = strongly agree). Need for affiliation. I adopted the IPIP AB5C: II+/I+ scale (DeYoung, Quilty, & Peterson, 2007) for warmth to measure the need for affiliation, or affiliative motivation. Affiliation is associated with warmth and social behaviors, therefore items like “Inquire about others’ well-being,” and “show my gratitude,” are acceptable measures of affiliation (Depue & Morroe-Strupinksy, 2005). The scale has 11 items (α = .91). Leaders rated how much they agree or disagree with each statement about themselves (1 = strongly disagree to 5 = strongly agree). Need for achievement. I adopted the IPIP NEO-C4 scale (Goldberg, 1995) for the need for achievement, or achievement motivation. It has 10 items (α = .86). An example item is “set 32 high standards for myself and others.” Leaders rated how much they agree or disagree with each statement about themselves (1 = very inaccurate disagree to 5 = very inaccurate). Incivility. I adapted Cortina, Magley, Williams, and Langhout’s (2001) scale for measuring workplace incivility. It has 7 items (α = .80). An example item is, “during the past week at work, have you been in a situation where you performed any of the following behaviors: ignored or excluded [Employee] from professional camaraderie?” Leaders reported how frequently they performed each behavior using a slider from 0 to 10+ times in the last week. Controls. I measured a leader’s gender, sex, age, education, tenure as a leader, organization size, number of direct reports, and the frequency of communication between leaders and followers. Past leadership studies have controlled for these variables (Bernerth, Cole, Taylor, & Walker, 2017). I elected to include them in this study because a leader’s demographic information has been shown to create meaningful differences in the workplace (e.g., differences in perception of leader success based on gender). Additionally, a leader’s tenure, organization size, number of direct reports, and frequency of communication will all affect how frequently they interact with employees (e.g., providing more or less opportunity to build a perceived sense of trust from followers). There were no differences in significance between models run with these control variables versus those without. I ran all models with both control variables and without them. For the sake of parsimony, I present the models without control variables. Attention Checks. Three attention check questions were included to detect acquiescent participant responses. However, no participants answered these questions incorrectly (e.g., “Select ‘strongly agree’ for this question,” or qualitative coding for realistic job titles for a leadership role). This is unsurprising as online survey populations with high completion rates of 33 HITs have been found to perform similarly with or without the presence of such questions (Keith, 2017). Analysis To test my hypotheses, I conducted a path analysis in R to test individual hypotheses between predictors and endogenous variables before testing the overall model in a holistic fashion with bootstrapping to test indirect effects. Missing data analyses were conducted in SPSS, and while no patterns emerged between demographic variables and missing data, four variables had roughly 10% missing data or more from their scale scores. Those variables were abusive supervision (n = 203, 19.4% missing), initiating structure (n =226, 10.3% missing), employee’s task performance (n = 218, 13.5% missing), and incivility (n = 228, 9.5% missing). Most participants responded to almost all items within a scale; however, some respondents chose not to respond to any items within the abusive supervision or initiating structure scales. This might be an effect of social desirability. While randomizing the survey by scale has been found to reduce social desirability responding in some meta-analyses (Lensvelt-Mulders et al., 2005), other studies have found that this is not always a solution, especially when dealing with sensitive questions in which participants might want to respond in a socially desirable way (Hoglinger & Diekmann, 2017). As participants might feel it would be most socially desirable to avoid responding poorly about themselves or their employees’ performance at work, it is plausible that some participants chose not to respond. This might be especially true given that this sample is largely based in Europe where data privacy and security laws are robust, potentially instilling mistrust when asked to share sensitive personal information online. Regardless of cause, the data are not missing completely at random (MCAR). Therefore, I used listwise deletion to remove these cases from my analyses (Enders & Bandalos, 2001). While there were outliers in the 34 dataset, removing them had little to no effect on the relationships between variables in this study. They were allowed to remain in the data set. Items were parceled for all scales in preparation for confirmatory factor analysis (CFA). Item parceling followed recommendations from Hall, Snell, and Foust (1999), such that theoretically related items were grouped together (e.g., trust had three parcels: ability, benevolence, and integrity, etc.). If not theoretically related, items that demonstrated similar content (e.g., tasks on the job vs interpersonal interactions in task performance) and loadings in an exploratory factor analysis (EFA) were parceled. If items did not share theoretical justification or content, the factor was given 3 parcels and items were assigned to parcels in pairs of highest and lowest loadings on the single latent factor (Table 2). Measurement models that include all the substantive variables were evaluated using confirmatory factor analyses (CFA). I evaluated CFA goodness of fit using the comparative fit index (CFI), the Tucker-Lewis index (TLI) and root mean square error approximation (RMSEA). CFI values at .95 or above, TFI values at .95 or above, and RMSEA less than .08 indicate good model fit (Hooper, Coughlan, & Mullen, 2008). To test overall model fit, I used a path model The Lavaan package for path analysis in R (Rosseel, 2012). While model fit statistics are available, they are less informative than evaluating the regression weights of the relationships between the variables in my model to better understand the relationship between variables. I report the unstandardized beta coefficient (B), standard error, z-value, and significance level for each significant relationship (Billings & Wroten, 1978). The significance of interaction terms’ beta weights was used to test moderation hypotheses. When a significant moderation effect was found, it was further analyzed using 35 simple slopes (Preacher, Curran, & Bauer, 2006). While not hypothesized, the model implies a moderated mediation model. To test the moderated mediation, I used the index of moderated mediation as proposed by Hayes (2015). This analysis is based on previous work on conditional process analysis and utilizes bootstrapped confidence intervals to make inferences about the data (e.g., if the confidence interval does not include 0, there is an effect). This analysis was be run using syntax in R based on Hayes’s (2015) process model (bootstrapping = 2000 iterations). The additional exploratory hypotheses for the moderating effects of leader and moral identities was analyzed using the same modeling technique as the currently proposed moderators. RESULTS Table 3 shows descriptive statistics and Table 4 shows correlations and reliabilities among study variables, as well as additional correlations among study variables and exploratory factors of trust. Confirmatory Factor Analysis (CFA) Confirmatory factor analyses (CFAs) were performed to examine the distinctiveness of the variables (employee task performance, employee voice behavior, leader’s perceived trust from follower, abusive supervision, initiating structure, consideration behavior, incivility, positive reciprocity orientation, need for power, need for achievement, and need for affiliation). As shown in Table 5, the 11-factor CFA model had satisfactory fit (Chi-squared = 628.92, df = 409, p < .001, CFI =.92, TLI = .91, RMSEA = .06, SRMR = .06) that was better than three 10- factor models, two 9-factor models, one 6-factor model, one 5-factor model, and one 4-factor model (Table 5). Hypotheses 36 To test my hypotheses, I conducted a path analysis in R (Rosseel, 2012). Each hypothesis was individually tested as proposed. All hypotheses were then tested together in a moderated mediation path model with bootstrapping to detect indirect effects. Missing data were handled using listwise deletion as participant responses were not missing at random. The same path models were also tested using full information maximum likelihood estimation. Many participants chose not to answer questions regarding abusive supervision. Of the control variables included in the study (number of direct reports, hours per week worked, interactions with followers per week, gender, sex, age, race, tenure, organization size, and virtual days worked per week), only the following had significant correlations with the study variable listed in parentheses: number of direct reports (leader’s perceived trust), interactions per week (leader’s initiating structure), age (ethical leadership). However, controlling for these variables did not change significant results. I report models without control variables for parsimony. The following reported regression weights for regression equations are unstandardized (B). Results of specific hypotheses are presented in the following paragraphs. Tables for each proposed hypothesis are presented in the Tables section. Results are interpreted as significant if p < .05; however, p < .10 are also interpreted in the results section, despite not being significant. Because of the relatively large number of individuals removed due to missing data in some scales, it is possible that some of the results that are approaching significance (p < .10) might have been significant had the final sample size for that model been larger. I first ran barebones path models without mediation or moderation effects before building more complex models by adding additional variables (e.g., mediators, moderators, and exploratory variables). I report results in tables grouped by predictors, outcome variables, mediation, and finally with moderation. All analyses were first tested in simple regressions. If there were no differences in the resulting 37 significant effects between simple regressions and path models without moderation or mediation, I omit reporting most simple regressions for parsimony. Results are summarized in a final, comprehensive table at the end of the results section. I first estimated a model for the main effects of my predictors (employee voice and task performance) to my mediator (leader’s perceived trust). Hypothesis 1 was supported, where employee task performance was positively related to a leader’s perceived trust from their follower (B = 0.06, s.e. = 0.02, Z = 3.06, p = 0.01). However, Hypothesis 2 was not supported, as voice behavior was not positively related to a leader’s perceived trust from employees (B = - 0.01, s.e. = 0.02, Ζ = -0.18, p = 0.86). Both employee task performance and voice behavior accounted for a small amount of variance in a leader’s perceived trust from their follower (R2 = 0.08). These results are summarized in Table 6. Hypothesis 3a was not supported, as positive reciprocity orientation did not moderate the relationship between employee task performance and a leader’s perceived trust from followers (B = 0.02, s.e. = 0.04, Ζ = 0.52, p = 0.65). These results are summarized in Table 8. There was a moderation of positive reciprocity orientation on the relationship between employee voice behavior and a leader’s perceived trust from followers (B = -0.06, s.e. = 0.03, Ζ = -1.98, p = 0.05). The simple slopes showed that the relationship between a follower’s voice behavior with leader’s perceived trust from the follower was weaker when the leader’s reciprocity orientation is high (B = -0.01, s.e. = 0.02, p < 0.44; see Table 8 and Figure 2) versus low (B = 0.06, s.e. = 0.02, p < 0.01). Despite the significant moderation, the simple slopes pattern was inconsistent with Hypothesis 3b, thus providing no support for the hypothesized moderation of positive reciprocity orientation on the relationship between a follower’s voice behavior with leader’s perceived trust from the follower. 38 Hypothesis 4 was not supported, as a leader’s perceived trust from a follower was not positively related his or her abusive supervision towards the follower (B = -0.05, s.e. = 0.04, Ζ = -1.30, p = 0.20). These results are summarized in a path model in Table 9 with the other originally proposed outcome variables (consideration behavior and initiating structure) and were the same as those produced simple regressions. Hypothesis 5 was not supported, as a leader’s need for power did not moderate the relationship between leader’s perceived trust from a follower with abusive supervision (B = - 0.03, s.e. = 0.05, Ζ = -0.65, p = 0.51). These results are summarized in Table 10. Hypothesis 6 was not supported, as a leader’s perceived trust from a follower was not negatively related to his or her initiating structure behavior towards the follower. Instead, a leader’s perceived trust from a follower was positively related to his or her initiating structure behavior (B = 0.66, s.e. = 0.25, Ζ = 2.65, p < 0.01). Perceived trust accounts for 5% of the variance in leader initiating structure. These results are summarized in Table 9. Hypothesis 7 was not supported, as a leader’s need for achievement did not moderate the relationship between leader’s perceived trust from a follower with his or her initiating structure (B = 0.67, s.e. = 0.40, Ζ = 1.69, p = 0.09). These results are summarized in Table 10. However, this result is approaching significance at p < .10, therefore I will interpret the moderation effect. The simple slopes showed that the relationship between a leader’s perceived trust with a leader’s initiating structure behavior was moderated by a leader’s need for achievement. This relationship is stronger when the leader’s positive reciprocity orientation was high (B = 0.71, s.e. = 0.36, p < 0.05) versus low (B = 0.37, s.e. = 0.34, p < 0.03) as can be seen in Table 11 and Figure 3. 39 Hypothesis 8 was not supported, as a leader’s perceived trust from a follower was not negatively related to his or her consideration behaviors towards the follower. Instead, a leader’s perceived trust from a follower was positively related to his or her consideration behavior (B = 0.81, s.e. = 0.24, Ζ = 3.33, p < 0.001). Perceived trust accounts for 6% of the variance in leader consideration behavior. These results are summarized in Table 90. Hypothesis 9 was not supported, as a leader’s need for affiliation did not moderate the relationship between leader’s perceived trust from a follower with consideration (B = -0.32, s.e. = 0.33, Ζ = -0.99, p = 0.32). These results are summarized in Table 10. Hypothesis 10a was not supported, as a leader’s perceived trust from a follower was not positively related to leader incivility towards the follower. Instead, a leader’s perceived trust from a follower was negatively related to leader incivility toward the follower (B = -0.20, s.e. = 0.07, Ζ = -2.84, p < 0.01). Perceived trust accounts for 4% of the variance in leader incivility behaviors. These results are summarized in Table 12. Hypothesis 10b was not supported, as a leader’s need for power did not moderate the relationship between leader’s perceived trust from a follower with incivility (B = -0.02, s.e. = 0.09, Ζ = -0.24, p = 0.81). However, there was a significant relationship between a leader’s need for power and their incivility (B = 0.10, s.e. = 0.04, Ζ = 2.35, p = 0.02), such that higher need for power was associated with more incivility. These results are summarized in Table 13. The implied moderated mediation was tested using R script based on Hayes’s (2015) process model. There were no significant moderated mediation effects in the model, as tested by the index of moderated mediation. This index compares the moderation effects on the indirect path before and after the mediator to see if there is a non-zero value when two moderators are 40 both present. All tested indices were not significant and their respective bootstrapped confidence intervals (iterations = 2000) contained 0. Exploratory Results Additional path models were examined with direct paths from the predictor variables (employee task performance and voice behavior) to the outcome variables (abusive supervision, initiating structure, and consideration behavior) to account for potential partial mediation effects. In this case, employee task performance was not significantly related to leader outcomes. Employee voice behavior had a positive relationship with initiating structure (B = 0.33, s.e. = 0.07, z = 4.99, p < 0.01) and consideration behavior (B = 0.30, s.e. = 0.05, z = 6.44, p < 0.01), and approached significance (p < .10) for abusive supervision (B = 0.02, s.e. = 0.01, z = 1.84, p = 0.07). Thus, employee voice behavior was positively associated with a leader’s abusive supervision, initiating structure, and consideration behavior (Tables 14-17). I then examined if positive reciprocity orientation moderated any of the direct paths between employee task performance and voice behavior and the outcome variables. Two moderation effects were significant at p < .05. The relationship between employee voice behavior with a leader’s initiating structure behavior was moderated by a leader’s positive reciprocity orientation (B = -0.25, s.e. = 0.10, z = -2.51, p = 0.01). This relationship is weaker when the leader’s positive reciprocity orientation was high (B = 0.20, s.e. = 0.06, p < 0.001) versus low (B =0.44, s.e. = 0.08, p < 0.001; see Table 18 and Figure 4). The relationship between employee task performance with a leader’s initiating structure behavior was moderated by a leader’s positive reciprocity orientation (B = 0.27, s.e. = 0.10, z = 2.62, p = 0.01). This relationship is stronger when the leader’s positive reciprocity orientation 41 was high (B = 0.30, s.e. = 0.10, p < 0.01) versus low (B =0.28, s.e. = 0.28, p < 0.01; see Table 19 and Figure 5). Exploratory Research Questions I added additional exploratory research questions in case moral licensing could not explain leader behaviors. If moral licensing could not, then perhaps idiosyncratic credits could. Leaders might leverage their perceived trust to behave in additional, positive ways (e.g., pushing their followers to change for the better, consider additional ethical quandaries, perform fewer bad leadership behaviors, etc.). For these reasons, I included ethical leadership and laissez-faire leadership as additional outcome variables on which idiosyncratic credits could be spent. Idiosyncratic credits were represented by leadership identity and moral identity as potential moderators of ethical leadership and laissez-faire leadership respectively. Ethical Leadership. To test the exploratory relationships between variables in my research model and the additional outcomes, I first examined the relationships between study variables and ethical leadership. The relationship between a leader’s perceived trust on ethical leadership is approaching significance at p < .10 (B = 0.51, s.e. = 0.31, Ζ = 1.64, p = 0.10) (Table 20). 14% of the variance in ethical leadership was explained (Table 20). Including first-stage (positive reciprocity orientation) and second-stage (leadership identity) moderators did not affect significant results. Leadership identity was positively related to ethical leadership (B = 0.94, s.e. = 0.23, Ζ =3.99, p < 0.01). Including direct paths from the predictors (employee task performance and voice behavior) to ethical leadership produced significant associations: employee task performance (B = 0.15, s.e. = 0.07, Ζ = 2.14, p < 0.03) and voice (B = 0.33, s.e. = 0.06, Ζ = 5.69, p < 0.01) both had a significant relationship with ethical leadership. Together, these predictors accounted for 37% of the variance in ethical leadership. Additionally, the 42 relationship between perceived trust and ethical leadership was approaching significance at p < .10 (B = 0.47, s.e. = 0.27, Ζ = 1.72, p = 0.08). Leadership identity did not moderate the relationship between perceived trust and ethical leadership (Tables 20-24). Laissez-faire leadership. The second set of exploratory analyses I ran focused on laissez-faire leadership as an outcome variable. A leader’s perceived trust from followers was not significantly related to a leader’s laissez-faire leadership behavior. There were no significant moderated mediation effects, and no direct paths were significant in additional models (Tables 25 and 26). A leader’s perceived trust was approaching significance when negatively related to laissez-faire leadership (B = -0.23, s.e. = 0.14, Ζ = -1.62, p = 0.10). “Everything but the kitchen sink” model. A full model including the original proposed hypotheses, ethical leadership, and laissez-faire leadership was tested using a path model in R. Additional moderators of leadership identity and moral identity were included for the two respective additional outcome variables, ethical leadership and laissez-faire leadership. Results were not different from those described in the analyses above. There were no new significant relationships. There were no significant moderated mediation effects in the model as tested by the index of moderated mediation. This index compares the moderation effects on the indirect path before and after the mediator to see if there is a non-zero value when two moderators are both present. All tested indices were not significant and their respective bootstrapped confidence intervals (iterations = 2000) contained 0. These results are summarized in tables 27-29. Trust as Ability, Benevolence, and Integrity. Additional exploratory path models were examined to treat perceived trust as three distinct factors (Mayer & Davis, 1995). The base structure of these path models was employee task performance and voice behavior as predictors for each of the leader outcome variables in the study (abusive supervision, initiating structure, 43 consideration, ethical leadership, and laissez-faire leadership). This relationship was estimated to be fully mediated by the three factors of trust: ability, benevolence, and integrity. I ran simple path models with one leadership outcome per model. They revealed unique relationships between the facets of trust and the employee predictors and leadership outcomes (Tables 30-35). Across these barebones models, employee task performance is always significantly related to a leader’s perceived integrity (B = 0.07, se = 0.03, z = 2.79, p = 0.01) (Table 35), sometimes the relationships approaches significance with perceived benevolence (B = 0.05, se = 0.03, z = 1.73, p < 0.08), and sometimes a significant (B = 0.05, se = 0.03, z = 1.94, p = 0.04; Table 32) or approaching significant relationship (B = 0.05, se = 0.03, z = 1.73, p = 0.08; Table 33) with perceived ability. Employee voice behavior is associated with different factors of trust differently. Employee voice had a significant negative relationship with perceived integrity one time (B = -0.04, se = 0.02, z= -1.99, p = 0.05; Table 32). When in the context of a larger path model (Table 36), the introduction of other variables results in only one significant relationship: that between employee task performance and perceived integrity (B = 0.07, se = 0.03, z = 2.58, p = 0.01). Employee voice behavior is approaching significance in its relationship with perceived benevolence (B = 0.04, se = 0.02, z = 2.58, p = 0.09) and perceived integrity (B = -0.04, se = 0.02, z = -1.73, p = 0.08). However, none of the three factors have a significant relationship with employee task performance once direct paths from task performance to outcome variables are included. The relationships between the three factors of trust and the subsequent leader behaviors also varied: perceived benevolence had a positive and significant relationship with ethical leadership (B = 0.42, s.e. = 0.21, Ζ = 2.03, p < 0.04; Table 36) in models without direct paths, and perceived benevolence was approaching significance in its relationship to laissez-faire 44 leadership (B = -0.19, s.e. = 0.12, Ζ = -1.68, p < 0.09; Table 36). However, both of these effects failed to persist when direct paths and the moderation of those direct paths were included. Regardless of the complexity of the model, perceived ability was positively related to a leader’s initiating structure (B = 0.51, s.e. = 0.24, Ζ = 2.17, p = 0.03; Table 37); and perceived benevolence was positively related to a leader’s consideration behavior (B = 0.85, s.e. = 0.18, Ζ = 4.62, p < 0.01; Table 37). As mentioned, once direct paths between predictor variables and outcomes were included in addition to the mediated pathways, the three factors of trust had fewer significant relationships with leader behaviors. The more complex the model, the less variance the three factors of trust explained. To summarize the path model in Table 36, a leader’s perceived employee perceptions of their ability, benevolence, and integrity are associated differently with both employee task performance, which is positively associated with all three factors (Table 30), and voice behavior, which does not approach significance until more variables are introduced into the path model. The three factors of trust are then related to different leader outcome variables: as perceived ability increases, initiating structure also increases (in simple models); as benevolence increases, both consideration and ethical leadership also increase; and lastly perceptions of integrity had no significant effects on the measured leader behaviors. Some of these relationships are subsumed by the introduction of other variables (Tables 37-39) such as first- and second-stage moderators, as well as moderated direct paths between predictor variables (task performance and voice) and leadership outcomes. Nonetheless, I have found some evidence that perceptions of the three facets of trust have different relationships with different leader outcomes. In testing these additional path models with the three facets of trust I found additional significant interaction effects. A leader’s positive reciprocity orientation moderated the 45 relationship between employee voice behavior and a leader’s perceived ability (B = -0.11, s.e. = 0.05, Ζ = -2.30, p = 0.02; Table 37). The simple slopes showed that the relationship between a follower’s voice behavior and the leader’s perceived ability was weaker when positive reciprocity orientation was high (B = -0.04, se = 0.02, p < 0.11) versus low (B = 0.07, se = 0.03, p = 0.01; see Table 40 and Figure 6). This result is inconsistent with the hypothesized direction for trust as a single construct. A moderation effect of a leader’s positive reciprocity orientation was approaching significant moderation on the relationship between employee voice behavior and a leader’s perceived integrity at the level of p < .10 (B = -0.08, s.e. = 0.04, Ζ = -1.81, p = 0.07). The simple slopes showed that the relationship between a follower’s voice behavior and the leader’s perceived integrity was not significant at high (B = -0.03, s.e. = 0.02, p = 0.24) or low (B = 0.03, s.e. = 0.03, p = 0.26) values of positive reciprocity orientation (see Table 41 and Figure 7). This was inconsistent with the hypothesized effect of positive reciprocity orientation on the relationship between employee voice behavior and trust as a single aggregate. Rather than see no effect, I had anticipated a stronger relationship as positive reciprocity orientation increased. Additionally, the moderation effect of a leader’s need for power on the relationship between a leader’s perception of their own benevolence and abusive supervision was significant (B = -0.06, s.e. = 0.03, Ζ = -1.94, p = 0.05). The simple slopes showed that the relationship between a leader’s perceived benevolence from their followers with their abusive supervision was stronger when their need for power was low (B = -0.11, s.e. = 0.03, p < 0.01) versus high (B = 0.01, s.e. = 0.03, p = 0.61; see Table 42 and Figure 8). This is antithetical my hypothesis on the moderation effects of the need for power on the relationship between trust as an aggregate and 46 abusive supervision. I had proposed that higher need for power would strengthen the relationship as opposed to weakening it. Alternate Model Estimation. The previous models used maximum likelihood (ML) estimation with listwise deletion of missing data. However, many of the relationships between variables were not significant (p > 0.05). To address potential issues due to omitting roughly two-fifths of responses through listwise deletion in ML for some models, additional models were run using full information maximum likelihood (FIML) estimation. FIML utilizes all available information, including partially complete cases, without imputing missing values and is superior across multiple conditions of missing data (Enders & Bandalos, 2009). There were no differences in significance levels for regression coefficients or indices of moderated mediation when comparing the two estimation methods. For parsimony, only the listwise deletion results are reported. Results Summary. Table 43 summarizes the results section. Hypotheses are addressed first, exploratory path models with additional outcomes second, and exploratory path models with three mediators third. DISCUSSION This study proposed to examine moral licensing in the context of the leader-follower relationship. Specifically, I proposed that leaders would accrue moral credits and credentials through social exchanges with followers in which they would then take subsequent license to perform fewer good behaviors and more negative behaviors based on leader-reported perceptions of perceived trust from their followers as moral credits or credentials. However, the results of this study provide limited support for these ideas. Instead, I found that the effect of leader perceptions of followers’ trust in them on the subsequent leader behaviors is often nonsignificant 47 and sometimes overshadowed by the direct effect of the leader’s perceptions of follower behaviors. The findings of this study support the conclusion that while leaders likely build perceptions of trust from their followers through repeated social exchanges, the data do not address my question of whether trust is spent as moral credits. I will first discuss the results of the front half of the model within the context social exchange theory. I will then discuss the leader outcomes in the context of moral licensing, the role that individual differences might play in this context, and support that trust should be measured as three distinct factors (ability, benevolence, and integrity). Theoretical Contributions My study found evidence that Social Exchange Theory (SET) can be used to explain how leader's build perceptions of feeling trusted by followers. SET argues that once trust is developed, it is more predictive of subsequent behaviors than more transactional measures (Dirks & Ferrin, 2002; MacKenzie, Moorman, & Fetter, 1990). This argument supports the finding that transactional behaviors such as employee task performance were positively related to a leader’s perception of feeling trusted. Therefore, while SET might explain the mechanism of the relationship between employee task performance and a leader’s perceived trust, it does not explain how voice behavior fails to make leaders feel more trusted. I argued that leaders might then use this perceived trust as moral credits or credentials to take license for future negative behaviors. This was largely unsupported by my data and limited my findings. Moral licensing theory states that good results eventually lead to poor outcomes (Miller & Effron, 2010) when an individual takes licenses through accumulated moral credits or credentials. Instead, I found evidence that leaders were most likely to choose good behavior consistent with their thoughts on their followers’ perceptions of them. There was little variance 48 in the number of negative behaviors leaders reported with most reporting low levels of abusive supervision and incivility. Had the model supported the theoretical perspective put forward by moral licensing, I would have expected to see leaders behaving more poorly when perceived trust was high—meaning negative relationships with good behaviors and positive with bad. Instead, a leader’s feeling of being trusted was positively related to their initiating structure, consideration, and ethical leadership behaviors, and negatively related to their incivility. There were no significant relationships between a leader’s perceived trust and abusive supervision or laissez-faire leadership. The lack of significant relationships with abusive supervision and laissez-faire leadership is likely because these leader behaviors occurred in such small numbers within my data. Comparing between leaders to understand when one leader might take license when others do not is not feasible when the negative behaviors I am investigating are reported in such low numbers within my data set. Therefore these results only partially address the hypotheses I proposed around moral licensing. Because of the positive significant relationships between perceived trust and some positive leader behaviors, there is likely little support that leaders take moral license to perform fewer good behaviors. Given the lack of variance in leaders’ performance of negative behaviors, I do not have enough data to assess if leaders take moral license to engage in more negative behaviors. Further confounding these interpretations is that perceived trust is being used as a proxy for moral credits or credentials in this study. It might have been helpful to include a variable to represent moral credits or credentials instead. This would eliminate what might be a missing variable problem and miss specified model. Perhaps there would have been differences in how perceived trust is spent as a credit compared to how moral license is taken. Without including this variable, this comparison, and thus a test of moral licensing theory, are not quite complete. 49 Additionally, I posited that certain leaders might take moral license as moderated by individual differences. As Miller and Effron note (2010), we do not yet understand why individuals differ in their decision to take moral license. My inclusion of first- (positive reciprocity orientation) and second-stage (needs for achievement, affiliation, and power; leadership identity; moral identity) moderators attempted to address this question. Positive reciprocity orientation moderated the relationships between employee voice behavior with a leader’s perceived trust, employee voice behavior and ethical leadership, employee voice behavior with a leader’s initiating structure, employee task performance and a leader’s initiating structure, and a leader’s perceived trust’s relationship with initiating structure. The results were mixed with higher levels of positive reciprocity orientation strengthening the relationship between perceived trust and initiating structure. However, higher levels of positive reciprocity orientation also weakened the relationship between employee voice and a leader’s initiating structure, and employee voice behavior and a leader’s perceived trust. The effect on the relationship between employee voice and a leader’s perceived trust contradicts what I anticipated in Hypothesis 3b; the strengthened relationships align with the theoretical perspective that informed Hypothesis 3b, yet they still do not support it directly. In short, these findings do not support my hypotheses based on moral licensing. One explanation might be that leaders respond to the intentions revealed by their employees’ actions as opposed to just the actions themselves when exhibiting positive reciprocity orientation (Ackerman, Fleiß, and Murphy, 2016). Leaders with lower levels of positive reciprocity orientation might assume better of their employees’ intentions than those with a higher positive reciprocity orientation. I did not measure perceived intentions of followers in my model though, so I cannot account for this possibility. This could also be related to employee voice behavior as a predictor in the relationships that were weakened 50 with higher reciprocity orientation. Leaders with higher levels of positive reciprocity orientation might have viewed increased voicing from employees as a breach of psychological contract (Ng, Feldman, and Butts, 2014). Voice behavior directed towards these leaders might lower perceptions of reciprocity as voice might breach of the leader-follower psychological contract for leaders with high positive reciprocity orientation (Wåhlin-Jacobsen, 2020). Future studies might investigate the context around assumed intentions of leader-follower pairs as well as perceived behaviors within the leader-follower relationship to better address this question. None of the other hypothesized moderation effects between a leader’s perceived trust with their subsequent behaviors were significant. Only when separating trust into its three factors as part of my exploratory analyses did I find a moderation effect of a leader’s need for power on the relationship between a leader’s perceived benevolence with their abusive supervision; and the moderation effect of reciprocity orientation on the relationship between employee voice behavior and a leader’s perceived ability, and employee voice behavior and a leader’s perceived integrity. The significant moderation effect of a need for power was antithetical to my proposed hypotheses despite not proposing a relationship between benevolence and abusive supervision. However, some studies have found that high levels of both need for power and benevolent leadership values produce better leadership outcomes despite their theoretically opposite definitions (Sverdlik, Oreg, and Berson, 2022). While I measured perceptions of benevolence instead of benevolent leadership values, this relationship between power and benevolence could still be important. Future studies should investigate this relationship before any conclusions are made. This is especially true given the low base rate of abusive supervision in this study. None of these exploratory moderation effects provide support for my hypothesis that leaders will spend their perceived trust as moral credit or credentials to perform more bad behaviors and fewer good 51 ones. These exploratory analyses instead suggest differences in how both employee task performance and voice behavior relate to the three facets of a leader’s perceived trust and other leader behaviors in my study. Task performance has a positive relationship with ability, benevolence, and integrity; and voice behavior has a negative relationship with integrity. Voice behavior has relationships that are opposite of those I predicted in my hypotheses with trust as an aggregate. In the latter half of the model with trust’s three factors as mediators, ability is positively related to initiating structure; integrity is negatively related to abusive supervision; and benevolence has positive relationships with both consideration and ethical leadership. These findings are consistent with previous literature which suggests that the three factors of trust will behave differently in different contexts (Butler and Cantrell, 1984), and also provide some support for the behavioral consistency perspective (e.g., good behavior begets more good behavior). While there were some significant moderations, there was no overall pattern in which leader identity or individual differences were perceived to have been instrumental in valuing trust as a form of credit that could later be spent—as an aggregate mediator or its three component factors. This might once again be due to the lack of a variable to explicitly measure moral credits or credentials. Individual differences had larger and more often significant effects when moderating the direct paths from employee behaviors to leader outcomes. While this does not support my hypotheses around individual leader traits moderating the relationships between trust and leader outcomes, it does support the broader argument that individual differences in leaders will change their behavior. Previous studies have found evidence for the effect of reciprocity orientation on leader behavior (Mitchell & Ambrose, 2007; Sverdlik, Oreg, & Berson, 2022; Vriend et al., 2020; Wei & Si, 2013) with specific evidence of positive reciprocity 52 orientation moderating leader behaviors (Umphress, Bingham, & Mitchell, 2010). Another possible explanation for the lack of moderation effects supporting moral licensing in my study might be because I used a between-subjects design. The paper by Lin, Ma, and Johnson (2016) used a within-person design which might have been more appropriate given that moral licensing is a within-person process. I posed an alternative theoretical perspective to moral licensing—that leaders might accrue idiosyncratic credits instead and continue to participate in good behavior in what the literature describes as the behavioral consistency effect (Zhu, May, & Avolio, 2004). This would have been supported had there been any significant moderation effects due to a leader’s leadership identity. However, there were none. Because leader identity was not a significant moderator in my model, using a leader’s perceived trust as any form of credit to drive subsequent behavior is largely unsupported. Indeed, if idiosyncratic credits were to serve as an alternate explanation, I would have expected to also see stronger positive relationships between how much trust a leader thought their employees had in them with their positive leader behaviors and weaker relationships with negative leader behaviors. Because this result was only seen at the bivariate level, it is likely that leaders accrue idiosyncratic credits instead of moral credits, and therefore they cannot use them to license different, potentially more positive behavior. Instead, it is likely that leaders choose to continue a series of good behaviors. Had I included a variable to operationalize moral credits in the current research, I might be able to speak to the expenditure of perceived trust compared moral credits or credentials. As such, these results can only be said to provide some support for the behavioral consistency effect. While this study found fewer significant relationships relating to a leader’s perceived trust than predicted, this is only one of four ways to interpret the trustor-trustee relationship (e.g., 53 Do I trust you? Do you trust me? Do I think you trust me? Do you think I trust you?). These four perspectives might be related to other important behaviors or outcomes that manifest over time. Therefore, future studies and applied work might consider how trust is not only multifaceted in its factor structure, but also how trust is multifaceted in its interpretation by both trustors and trustees. Studies on trust have largely focused on the trustor—asking if followers (the trustors) trust their leader. However, trust is a reciprocal interaction in which both parties trust each other (e.g., dyadic trust). Both halves of the dyad think about if the other side trusts them or not, and the literature has paid less attention to these perceptions (Tepper, Henle, Lambert, Giacalone, & Duffy, 2008; Podsakoff, Todor, & Skov, 1982; Korman, 1966; Bass & Stogdill, 1990). To summarize, my study provides partial support that perceived trust has significant relationships with leader reports of employee behaviors, and secondly, that while there were some small significant associations with feeling trusted by an individual, the direct relationships between employee behaviors and leader outcomes were much stronger. Individual differences have some relationships with the variables in this model, but not as many as predicted. The largest contribution is that this study examines trust from the perspective of the trustee, rather than the trustor. While this is interesting to consider, the relationships stemming from the trustee’s reflection on how trusted they might be are relatively unsupported. Practical Implications This study demonstrates that different components of perceptions of trust (ability, benevolence, and integrity) predict different leader behaviors. Organizations might benefit from considering how these different relationships inform their internal frameworks for culture, learning and development, as well as diversity, equity, inclusion, and belonging initiatives. For example, if leaders perceive themselves to possess ability in the eyes of their followers, they are 54 less likely to engage in initiating structure. An organization seeking to increase initiating structure might then train employees and leaders on behaviors associated with feeling capable, and what it means to be capable in their respective roles, in the workplace. The same goes for perceptions benevolence to consideration and ethical leadership. Lastly, and most broadly, this study demonstrates that leaders’ perceptions of being trusted by their followers might add some small incremental value to creating a good environment for employees. Therefore, there is value in fostering these perceptions. Organizations and leaders should prioritize developing trust and leveraging resources to foster and develop interpersonal relationships across levels of the organizational hierarchy for both employees and leaders. However, perceptions of being trusted, while potentially helpful in specific situations, are not the most important aspect of trust in workplace relationships. Limitations My research is marked by several limitations. First, this study relied on a single-source, monomethod approach by using leader self-reports of their leadership behaviors, perceptions of follower behaviors, and individual differences. While participants answered questions across multiple time points, there is a concern that merely self-reporting about behaviors and attitudes can result in spurious relationships between variables because of their shared common measurement method. This is known as common method variance (CMV) and has, in some past studies, likely been the reason that two historically unrelated constructs have a significant relationship (Lindell & Whitney, 2001). Perhaps some of the counterintuitive findings in this study (e.g., employee voice behavior associated with abusive supervision) are due to common method variance. Further studies should be conducted using different methods to address this question. 55 Secondly, participants often engage in socially desirable responding when answering questions about poor or negative behaviors as part of either self-deception or impression management (Zerbve & Paulhus, 1987). Given that roughly 20% of participants chose not to answer questions about abusive supervision, a construct comprised of negative interpersonal behaviors, it is likely that participants were attempting to respond in a socially desirable way. This can be a problem because it creates patterns in missing data that are not at random, but also not related to any other underlying variable. Compounding this challenge is the low frequency with which leaders are likely to engage in abusive supervision. Because of its low base rate, reporting instances of abusive supervision over the span of 15 days might not capture any incidences when using behavior-based measures (e.g., “How many times in the past week did you… [answer using a slider between 0 and 10+]”). Together, these two limitations constrain the inferences I can make about negative, low base rate phenomena within this study. Thirdly, the sample size required to have enough power to find significant relationships was potentially low given the complexity of the path models analyzed here (Westland, 2010; Iacobucci, 2009). While all of my path models converged thanks to sample size recommendations from Preacher and Koffman (2006), the exploratory analyses were increasingly complex. The number of estimated paths nearly tripled when the mediator, trust, was split into its three factors. When estimating models fully mediated by the three factors of trust with both front- and back-end moderators, the number of paths being estimated exceeded 25. Therefore, there’s a chance that some relationships are spurious. To account for this, my interpretations considered whether the constructs were known to be related in previous studies, had theoretical justification for their relationship, and were significant in smaller, barebones 56 models. Future studies should consider recruiting more participants or reducing the number of constructs included in the models. Lastly, the number of models estimated when searching for significant relationships between proposed variables suggests cryptic multiple hypothesis testing. Cryptic multiple hypothesis testing means that because more models than were originally proposed were tested, the probably of finding “at least one significant effect is high, even if all null hypotheses are true” (Forstmeier & Schielzeth, 2011). While I report non-significant paths in my full model results, a lack of p value adjustments might have produced unidentified type I errors. This limits the reproducibility of my results in future studies and my ability to properly test the theories I set out to. CONCLUSION The perception of being trusted by followers has significant relationships with future leader behaviors, but ultimately actual employee behaviors have more significant associations with leadership behaviors. Leaders can consider why they feel trusted and how they might be able to leverage that trust (based on their perceived ability, benevolence, and integrity) to improve their employee’s lived experience through more positive leadership behaviors, but this is no substitute for developing trust (not just a perception of it). 57 Tables Table 1 Survey administration schedule. Surveys opened at the same time (GMT) regardless of the participant’s geographic location. Week 1 Week 2 Week 3 Survey 1 (Monday, Day 1) Survey 2 (Monday, Day 8) Survey 3 (Monday, Day 15) 58 Table 2 Number of parcels per scale. Scale Parcels Items per Parcel Employee Task Performance* 3 3, 2, 2 Employee Voice Behavior 3 2, 2, 1 Positive Reciprocity Orientation 3 4, 3, 3 Leader’s Perceived Trust from Followers* 3 6, 5, 6 Need for Achievement 3 4, 3, 3 Need for Affiliation 3 4, 4, 3 Need for Power 3 2, 2, 2 Initiating Structure 3 6, 5, 4 Consideration Behavior 3 6, 5, 4 Abusive Supervision 3 6, 5, 4 Incivility 3 2, 2, 2 Laissez-Faire Leadership Behavior 2 2, 2 Ethical Leadership 3 2, 2, 2 Moral Identity* 2 7, 5 Leadership Identity* 4 4, 4, 3, 4 Note: parcels with * were grouped by theoretical content. 59 Table 3 Descriptive Statistics and Correlations for Study Variables. T1, T2, and T3 represent measurement at times 1, 2, and 3 respectively. Variable n M SD 1. Employee Voice (T1) 245 3.56 2.41 2. Employee Task 237 7.07 1.87 Performance (T1) 3. Positive Reciprocity 246 3.89 0.61 Orientation (T1) 4. Trust (T2) 247 4.05 0.52 5. Laissez-faire 236 0.54 0.88 Leadership (T3) 6. Incivility (T3) 228 0.20 0.47 7. Ethical Leadership 232 3.59 1.91 (T3) 8. Consideration 229 5 1.65 Behavior (T3) 9. Initiating Structure 226 2.65 1.89 (T3) 10. Abusive Supervision 203 0.09 0.23 (T3) 60 Table 3 (cont’d) Variable n M SD 11. Need for Power (T1) 252 2.45 0.80 12. Need for Affiliation 246 4.16 0.58 (T1) 13. Need for 249 4.14 0.55 Achievement (T1) 14. Moral Identity (T1) 251 3.52 0.63 15. Leadership Identity 244 3.84 0.60 (T1) 16. Trust Factor: Ability 251 3.79 0.69 (T2) 17. Trust Factor: 250 3.71 0.75 Benevolence (T2) 18. Trust Factor: 249 3.60 0.67 Integrity (T2) 61 Table 4 Correlations and Reliabilities (Cronbach’s alpha (α); in parentheses on diagonals) for Study Variables. T1, T2, and T3 represent measurement at time 1, 2, and 3 respectively. Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1. Employee Task Performance (T1) (.81) 2. Employee Voice (T1) 0.39* (.86) 3. Positive Reciprocity Orientation 0.22* 0.06 (.85) (T1) 4. Trust (T2) 0.25* 0.10 0.17* (.92) 5. Laissez-faire Leadership (T3) 0.02 0.11 0.04 -0.10 (.76) 6. Incivility (T3) -0.06 -0.04 0.03 -0.18* 0.42* (.80) 7. Ethical Leadership (T3) 0.33* 0.50* 0.02 0.23* 0.15* 0.02 (.77) 62 Table 4 (cont’d) 8. Consideration Behavior (T3) 0.35* 0.60* 0.04 0.29* 0.08 -0.06 0.75* (.86) 9. Initiating Structure (T3) 0.29* 0.38* 0.06 0.20* 0.17* 0.12 0.62* 0.53* (.92) 10. Abusive Supervision (T3) -0.11 0.13 0.01 -0.09 0.51 0.54* 0.15* 0.06 0.24* (.77) 11. Need for Power (T1) -0.12 -0.05 0.07 -0.04 0.12 0.17* -0.08 -0.05 0.08 0.21* (.85) 12. Need for Affiliation (T1) 0.22* 0.18* 0.26* 0.44* -0.12 -0.19* 0.24* 0.29* 0.15* -0.06 -0.05 (.90) 13. Need for Achievement (T1) 0.20* 0.10 0.24* 0.44* -0.05 -0.10 0.17* 0.16* 0.20* -0.01 0.10 0.52 (.83) 14. Moral Identity (T1) 0.10 0.10 0.27* 0.17 0.01 -0.06 0.10 0.14* 0.07 0.01 0.02 0.43 0.33 (.87) 15. Leadership Identity (T1) 0.17* 0.31* 0.22* 0.43* -0.04 -0.05 0.34* 0.36* 0.20* 0.07 0.00 0.59 0.52 0.39 (.93) 16. Ability (T2) 0.14* 0.03 0.11 0.79* -0.06 -0.15* 0.21* 0.20* 0.22 -0.04 0.04 0.39* 0.38* 0.14* 0.33* (.91) 63 Table 4 (cont’d) - 17. Benevolence (T2) 0.19* 0.12* 0.15* 0.79* -0.08 0.19* 0.24* 0.33* 0.12 -0.17* -0.11 0.38*0.27*0.19*0.46*0.54* (.85) - 18. Integrity (T2) 0.21* 0.00 0.16* 0.74* -0.11 0.22* 0.18* 0.18* 0.15* -0.15* -0.10 0.29*0.33*0.11*0.24*0.51*0.51*(.69) Note: M and SD represent mean and standard deviation respectively. * p < .05. Reliabilities (α) presented in parentheses along the diagonal. T1, T2, and T3 represent measurement at time 1, 2, and 3 respectively. 64 Table 5 CFA results for parceled items. CFA Model Tested χ² df CFI TLI RMSEA SRMR 11-factor, full model 627.46* 409 0.92 0.91 0.06 0.06 10-factor: individual differences as 1 factor 922.58* 496 0.82 0.80 0.09 0.09 10-factor: voice and task perf. as one factor 669.19* 419 0.9 0.88 0.07 0.07 10-factor: negative leader behavior factor 776.31* 419 0.87 0.85 0.07 0.07 9-factor: grouped voice, task perf., and trust 837.15* 428 0.86 0.83 0.08 0.09 9-factor: positive and negative leader behaviors 971.98* 428 0.81 0.78 0.09 0.08 6-factor: positive and negative leader behaviors, individual differences, predictors, and the mediator 1444.71* 449 0.65 0.61 0.12 0.11 5-factor: positive and negative leader behaviors, individual differences, predictors, mediator 1498.52* 454 0.63 0.6 0.12 0.11 4-factor: leader behaviors, predictors, mediator, and individual differences 1827.81* 458 0.51 0.47 0.14 0.14 Note: * indicates p < .05. N = 163 of a possible 252 for the CFA models. Lavaan does not allow missing data in CFA, thus some participants were dropped. 65 Table 6 H1 & H2: Leader's perceived trust predicted by follower task performance and voice behavior. Regressions B s.e. Ζ-value p Perceived Trust Employee task performance 0.06* 0.02 3.06 0.01 Employee voice -0.01 0.02 -0.18 0.86 R-squared Perceived Trust 0.08 Note: * indicates p < .05. Table 7 Leader's perceived trust predicted by follower task performance and voice behavior. Both paths moderated by positive reciprocity orientation. Regressions B s.e. Ζ-value p Perceived Trust Employee task performance 0.05* 0.19 2.49 0.01 Employee voice 0.01 0.02 0.38 0.79 Positive reciprocity orientation 0.06 0.06 1.10 0.27 Task * pos. reciprocity ori. 0.02 0.04 0.52 0.65 Voice * pos. reciprocity ori. -0.06* 0.03 -1.98 0.05 R-squared Perceived Trust 0.13 Note: * indicates p < .05. Table 8 Simple slopes analysis for positive reciprocity orientation (mod), employee voice (x), and leader's perceived trust (y). See also: Figure 2. Slope S.E. t-value p-value SD +1 -0.01 0.02 -0.78 0.44 Mean 0.03* 0.01 1.94 0.05 SD - 1 0.06 0.02 3.32 < 0.001 Note: * indicates p < .05. 66 Table 9 H4, H6, H8: Results from a single path model of leader behaviors as predicted by a leader's perceived trust. Results are consistent with simple regressions from perceived trust to each outcome. Regressions B s.e. Ζ-value p Consideration Behavior Perceived trust 0.81* 0.24 3.33 0.00 Initiating Structure Perceived trust 0.66* 0.25 2.65 0.01 Abusive Supervision Perceived trust -0.05 0.04 -1.30 0.20 R-squared Consideration Behavior 0.06 Initiating Structure 0.05 Abusive Supervision 0.01 Note: * indicates p < .05. 67 Table 10 H5, H7, H9: Results from a single path model of leader behaviors moderated by motivation orientations for hypotheses 5, 7, and 9 as predicted by a leader’s perceived trust. Regressions B s.e. Ζ-value p Consideration Behavior Perceived trust 0.55* 0.29 1.93 0.05 Need for affiliation 0.42† 0.26 1.64 0.10 Perceived trust * Need for affiliation -0.32 0.33 -0.99 0.32 Initiating Structure Perceived trust 0.59* 0.30 1.97 0.05 Need for achievement 0.34 0.24 1.44 0.15 Perceived trust * Need for achievement 0.67† 0.40 1.69 0.09 Abusive Supervision Perceived trust -0.04 0.04 -1.00 0.32 Need for power 0.06* 0.02 2.57 0.01 Perceived trust * Need for power -0.03 0.05 -0.65 0.51 R-squared Consideration Behavior 0.09 Initiating Structure 0.06 Abusive Supervision 0.05 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). Table 11 Simple slopes analysis for positive reciprocity orientation (mod), a leader’s perceived trust (x), and initiating structure (y). See also: Figure 3. Slope S.E. t-value p-value SD +1 0.71* 0.36 1.99 0.05 Mean 0.54 0.28 1.92 0.06 SD - 1 0.37 0.34 1.11 0.27 Note: * indicates p < .05. 68 Table 12 H10a: Path model with three leader behaviors as outcomes for exploratory hypothesis 10a replacing abusive supervision with incivility. Regressions B s.e. Ζ-value p Consideration Behavior Perceived trust 0.88* 0.23 3.80 0.00 Initiating Structure Perceived trust 0.78* 0.26 2.98 0.00 Incivility Perceived trust -0.20* 0.07 -2.84 0.00 R-Squared Consideration Behavior 0.07 Initiating Structure 0.05 Incivility 0.04 Note: * indicates p < .05. 69 Table 13 H10b: Exploratory hypothesis 10b replacing abusive supervision with incivility, moderated by need for power. Regressions B s.e. Ζ-value p Consideration Behavior Perceived trust 0.62* 0.28 2.23 0.03 Need for affiliation 0.44† 0.24 1.80 0.07 Perceived trust * Need for affiliation -0.31 0.33 -0.95 0.34 Initiating Structure Perceived trust 0.71* 0.30 2.37 0.02 Need for achievement 0.39 0.24 1.60 0.11 Perceived trust * Need for achievement 0.63† 0.37 1.70 0.09 Incivility Perceived trust -0.19* 0.08 -2.35 0.02 Need for power 0.10* 0.04 2.35 0.02 Perceived trust * Need for power -0.02 0.09 -0.24 0.81 R-Squared Consideration Behavior 0.10 Initiating Structure 0.08 Incivility 0.06 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 70 Table 14 Proposed model with direct paths from predictors to leadership outcomes. Regressions B s.e. Z-value p Perceived Trust Employee task performance 0.05* 0.02 2.45 0.01 Employee voice 0.01 0.02 0.55 0.58 Positive reciprocity orientation 0.09 0.07 1.23 0.22 Task * pos. reciprocity ori. -0.01 0.04 0.12 0.90 - Voice * pos. reciprocity ori. 0.07† 0.04 -1.90 0.06 Consideration Behavior Perceived trust 0.48* 0.24 1.98 0.05 Need for affiliation 0.09 0.21 0.43 0.67 Perceived trust * Need for affiliation 0.01 0.27 0.05 0.96 Employee task performance 0.07 0.06 1.11 0.27 Employee voice 0.38* 0.04 9.46 0.00 Initiating Structure Perceived trust 0.66* 0.30 2.20 0.03 Need for achievement 0.25 0.25 1.10 0.31 Perceived trust * Need for achievement 0.65† 0.36 1.82 0.07 Employee task performance 0.06 0.08 0.70 0.48 Employee voice 0.28* 0.07 4.29 0.00 Abusive Supervision Perceived trust -0.18 0.03 -0.54 0.59 Need for power 0.05* 0.02 1.96 0.05 Perceived trust * Need for power -0.01 0.05 -0.26 0.80 Employee task performance -0.02 0.02 -1.33 0.18 Employee voice 0.02* 0.01 1.97 0.05 R-Squared Perceived Trust 0.16 Consideration Behavior 0.44 Initiating Structure 0.27 Abusive Supervision 0.09 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 71 Table 15 Proposed model with moderated direct paths from predictors to leadership outcomes. Regressions B s.e. Z-value p Perceived Trust Employee task performance 0.05* 0.02 2.53 0.01 Employee voice 0.01 0.02 0.57 0.57 Positive reciprocity orientation 0.09 0.07 1.30 0.20 Task * pos. reciprocity ori. -0.01 0.04 -0.36 0.72 Voice * pos. reciprocity ori. -0.08* 0.04 -1.94 0.05 Consideration Behavior Perceived trust 0.47† 0.25 1.87 0.06 Need for affiliation 0.25 0.21 1.19 0.24 Perceived trust * Need for affiliation 0.00 0.27 0.01 0.99 Employee task performance 0.06 0.06 0.99 0.32 Employee voice 0.30* 0.05 6.44 0.00 Task * pos. reciprocity ori. 0.09 0.10 0.92 0.36 Voice * pos. reciprocity ori. -0.05 0.07 -0.75 0.45 Initiating Structure Perceived trust 0.59† 0.31 1.90 0.06 Need for achievement 0.15 0.25 0.58 0.56 Perceived trust * Need for achievement 0.55 0.36 1.52 0.13 Employee task performance 0.04 0.08 0.46 0.64 Employee voice 0.33* 0.07 4.99 0.00 Task * pos. reciprocity ori. 0.27* 0.10 2.62 0.01 Voice * pos. reciprocity ori. -0.25* 0.10 -2.51 0.01 Abusive Supervision Perceived trust -0.02 0.03 -0.54 0.59 Need for power 0.04† 0.02 1.80 0.07 Perceived trust * Need for power 0.00 0.05 -0.08 0.93 Employee task performance -0.03 0.02 -1.36 0.18 Employee voice 0.02† 0.01 1.84 0.07 Task * pos. reciprocity ori. 0.02 0.02 0.90 0.37 Voice * pos. reciprocity ori. -0.01 0.02 -0.70 0.51 R-Squared Perceived Trust 0.15 Consideration Behavior 0.42 Initiating Structure 0.30 Abusive Supervision 0.09 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 72 Table 16 Indices of moderated mediation for the proposed path model with moderated direct paths from predictors to outcomes. Indices of Moderated Mediation Predictor Moderator (1) Mediator Moderator (2) Outcome B s.e. Z-value p ci.lower ci.upper Pos. Recip. Need for Voice Orientation Trust Affiliation Consideration 0.00 0.02 0.02 0.98 -0.06 0.04 Pos. Recip. Need for Task Orientation Trust Affiliation Consideration 0.00 0.01 0.00 1.00 -0.02 0.03 Pos. Recip. Need for Initiating Voice Orientation Trust Achievement Structure -0.05 0.04 -1.16 0.25 -0.17 0.01 Pos. Recip. Need for Initiating Task Orientation Trust Achievement Structure 0.00 0.03 0.09 0.92 -0.07 0.08 Pos. Recip. Abusive Voice Orientation Trust Need for Power Supervision 0.00 0.00 0.30 0.76 -0.01 0.01 Pos. Recip. Abusive Task Orientation Trust Need for Power Supervision 0.00 0.00 -0.03 0.97 -0.01 0.00 73 Table 17 Path model fit indices for an exploratory model with moderated direct paths from predictors to outcomes (Table 16). χ² df CFI TLI RMSEA SRMR 226.96* 69.00 0.57 0.38 0.12 0.10 Note: * indicates p < .05 Table 18 Simple slopes analysis for positive reciprocity orientation (mod), employee voice (x), and initiating structure (y). See also: Figure 4. Slope S.E. t-value p-value SD +1 0.20* 0.06 3.02 < 0.01 Mean 0.32* 0.05 6.33 < 0.01 SD - 1 0.44* 0.08 5.84 < 0.01 Note: * indicates p < .05. Table 19 Simple slopes analysis for positive reciprocity orientation (mod), employee task performance (x), and initiating structure (y). See also: Figure 5. Slope S.E. t-value p-value SD +1 0.30* 0.10 3.16 0.00 Mean 0.29* 0.07 4.18 0.00 SD - 1 0.28* 0.28 2.90 0.00 Note: * indicates p < .05. 74 Table 20 Ethical leadership path model with moderators. Regressions B s.e. Z-value p Perceived Trust Employee task performance 0.03 0.02 1.61 0.11 Employee voice 0.01 0.02 0.50 0.62 Positive reciprocity orientation 0.06 0.06 1.03 0.30 Task * pos. reciprocity ori. 0.01 0.04 0.21 0.83 Voice * pos. reciprocity ori. -0.06† 0.03 -1.92 0.06 Ethical Leadership Perceived trust 0.51† 0.31 1.64 0.10 Leadership identity 0.94* 0.23 3.99 0.00 Perceived trust * Leadership identity -0.29 0.33 -0.86 0.39 R-Squared Perceived Trust 0.10 Ethical Leadership 0.14 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 75 Table 21 Indices of moderated mediation for a path model of ethical leadership (Table 20). Variables B s.e. Z-value p ci.lower ci.upper Moderator Predictor (1) Mediator Moderator (2) Outcome Pos. Recip. Leadership Ethical Voice Orientation Trust Identity Leadership 0.02 0.02 0.70 0.48 -0.01 0.07 Pos. Recip. Leadership Ethical Task Orientation Trust Identity Leadership 0.00 0.02 -0.13 0.89 -0.04 0.02 76 Table 22 Path model for ethical leadership with moderated direct paths from predictors (employee task performance and voice behavior). Regressions B s.e. Z-value p Perceived Trust Employee task performance 0.03 0.02 1.62 0.11 Employee voice 0.01 0.02 0.49 0.63 Positive reciprocity orientation 0.06 0.06 1.03 0.30 Task * pos. reciprocity ori. 0.01 0.04 0.21 0.83 - Voice * pos. reciprocity ori. 0.06† 0.03 -1.88 0.06 Ethical Leadership Perceived trust 0.47† 0.27 1.72 0.08 Leadership identity 0.49* 0.19 2.52 0.01 Perceived trust * Leadership identity -0.02 0.30 -0.06 0.95 Employee task performance 0.15* 0.07 2.14 0.03 Employee voice 0.33* 0.06 5.69 0.00 Task * pos. reciprocity ori. -0.07 0.09 -0.79 0.43 Voice * pos. reciprocity ori. -0.01 0.09 -0.15 0.88 R-Squared Perceived Trust 0.10 Ethical Leadership 0.37 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 77 Table 23 Indices of moderated mediation for a simple path model of ethical leadership (Table 22). Z- Variables B s.e. value p ci.lower ci.upper Predictor Moderator (1) Mediator Moderator (2) Outcome Pos. Recip. Leadership Ethical Voice Orientation Trust Identity Leadership 0.00 0.02 0.06 0.95 -0.04 0.04 Pos. Recip. Leadership Ethical Task Orientation Trust Identity Leadership 0.00 0.01 -0.01 0.99 -0.03 0.02 78 Table 24 Barebones path model of predictors' (voice and task performance) and trust's relationships with ethical leadership. Regressions B s.e. Z-value p Perceived Trust Employee task performance 0.04* 0.02 2.24 0.02 Employee voice 0.00 0.02 -0.05 0.96 Ethical Leadership Perceived trust 1.01* 0.30 3.35 0.00 R-Squared Perceived Trust 0.05 Ethical Leadership 0.09 Note: * indicates p < .05. Table 25 Barebones path model of predictors' (voice and task performance) and trust's relationships with laissez-faire leadership. Regressions B s.e. Z-value p Perceived Trust Employee task performance 0.05* 0.02 2.61 0.01 Employee voice 0.00 0.02 -0.08 0.94 Laissez-faire Leadership Perceived trust -0.21 0.15 -1.42 0.16 R-Squared Perceived Trust 0.05 Laissez-faire Leadership 0.05 Note: * indicates p < .05. 79 Table 26 Path model for laissez-faire leadership with moderated direct paths from predictors (employee task performance and voice behavior). Only the path model with moderated direct paths is included here. Regressions B s.e. Z-value p Perceived Trust Employee task performance 0.04* 0.02 2.1 0.04 Employee voice 0.01 0.02 0.32 0.75 Positive reciprocity orientation 0.06 0.06 1.01 0.31 Task * pos. reciprocity ori. 0.01 0.03 0.26 0.80 Voice * pos. reciprocity ori. -0.06† 0.03 -1.86 0.06 Laissez-faire Leadership Perceived trust -0.23† 0.14 -1.62 0.10 Moral identity 0.03 0.08 0.38 0.70 Perceived trust * Moral identity -0.08 0.17 -0.47 0.64 Employee task performance -0.01 0.05 -0.15 0.88 Employee voice 0.05 0.04 1.29 0.20 Task * pos. reciprocity ori. 0.10 0.06 1.57 0.12 Voice * pos. reciprocity ori. -0.06 0.05 -1.08 0.28 R-Squared Perceived Trust 0.11 Laissez-faire Leadership 0.08 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 80 Table 27 Exploratory path model with direct paths from predictors to outcome variables. Regressions B s.e. Ζ-value p Perceived Trust Employee task performance 0.05* 0.02 1.98 0.05 Employee voice 0.01 0.02 0.49 0.63 Positive reciprocity orientation 0.08 0.07 1.08 0.28 Task * pos. reciprocity ori. -0.02 0.04 -0.48 0.63 - Voice * pos. reciprocity ori. 0.07† 0.04 -1.89 0.06 Consideration Behavior Perceived trust 0.61* 0.22 2.71 0.01 Need for affiliation 0.18 0.18 1.01 0.31 Perceived trust * Need for affiliation -0.05 0.23 -0.21 0.83 Employee task performance 0.09 0.06 1.54 0.12 Employee voice 0.38* 0.04 9.39 0.00 Initiating Structure Perceived trust 0.65* 0.31 2.13 0.03 Need for achievement 0.18 0.23 0.79 0.43 Perceived trust * Need for achievement 0.67* 0.34 1.98 0.05 Employee task performance 0.08 0.08 1.03 0.30 Employee voice 0.26* 0.07 3.93 0.00 Abusive Supervision Perceived trust -0.03 0.04 -0.85 0.39 Need for power 0.03 0.02 1.36 0.17 Perceived trust * Need for power -0.05 0.05 -1.07 0.28 Employee task performance -0.01 0.01 -1.40 0.16 Employee voice 0.01† 0.01 1.91 0.06 Laissez-faire Leadership Perceived trust -0.11 0.14 -0.78 0.44 Moral identity 0.02 0.07 0.29 0.77 Perceived trust * Moral identity -0.13 0.12 -1.10 0.27 Employee task performance 0.04 0.03 1.19 0.23 Employee voice 0.02 0.03 0.81 0.42 81 Table 27 (cont’d) Ethical Leadership Perceived trust 0.55† 0.32 1.70 0.09 Leadership identity 0.32† 0.20 1.64 0.10 Perceived trust * Leadership identity 0.15 0.30 0.50 0.62 Employee task performance 0.15† 0.08 1.81 0.07 Employee voice 0.33* 0.06 5.25 0.00 R-Squared Perceived Trust 0.13 Consideration Behavior 0.43 Initiating Structure 0.24 Abusive Supervision 0.05 Laissez-faire Leadership 0.03 Ethical Leadership 0.30 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 82 Table 28 Indices of moderated mediation for the exploratory path model with direct paths from predictors to outcomes (Table 28). Indices of Moderated Mediation Ζ- Predictor Moderator (1) Mediator Moderator (2) Outcome B s.e. value p ci.lower ci.upper Pos. Recip. Need for Voice Orientation Trust Affiliation Consideration 0.00 0.02 0.20 0.85 -0.03 0.04 Pos. Recip. Need for Task Orientation Trust Affiliation Consideration 0.00 0.01 0.09 0.93 -0.02 0.02 Pos. Recip. Need for Initiating - Voice Orientation Trust Achievement Structure 0.05 0.04 -1.23 0.22 -0.14 0.00 Pos. Recip. Need for Initiating - Task Orientation Trust Achievement Structure 0.01 0.03 -0.39 0.70 -0.08 0.05 Pos. Recip. Abusive Voice Orientation Trust Need for Power Supervision 0.00 0.00 0.98 0.33 0.00 0.01 Pos. Recip. Abusive Task Orientation Trust Need for Power Supervision 0.00 0.00 0.33 0.74 0.00 0.01 Pos. Recip. Laissez-faire Voice Orientation Trust Moral Identity Leadership 0.01 0.01 0.93 0.35 -0.01 0.03 Pos. Recip. Laissez-faire Task Orientation Trust Moral Identity Leadership 0.00 0.01 0.33 0.74 -0.01 0.02 Pos. Recip. Leadership Ethical - Voice Orientation Trust Identity Leadership 0.01 0.02 -0.51 0.61 -0.05 0.03 Pos. Recip. Leadership Ethical Task Orientation Trust Identity Leadership 0.00 0.01 -0.19 0.85 -0.04 0.02 83 Table 29 Path model fit indices (Table 28). χ² df CFI TLI RMSEA SRMR 479.27* 167.00 0.54 0.43 0.11 0.12 84 Table 30 Path model for trust as three factors including the direct paths from predictors to leader behaviors. Regressions B s.e. Z-value p Perceived Ability Employee task performance 0.05* 0.03 1.94 0.05 Employee voice -0.01 0.02 -0.25 0.80 Perceived Benevolence Employee task performance 0.05† 0.03 1.73 0.08 Employee voice 0.03 0.02 1.18 0.24 Perceived Integrity Employee task performance 0.08* 0.02 3.24 0.00 Employee voice -0.03 0.02 -1.29 0.20 R-Squared Perceived Ability 0.02 Perceived Benevolence 0.04 Perceived Integrity 0.05 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 85 Table 31 Path model for trust as three factors mediating the relationship between employee task performance and voice behavior with abusive supervision. Regressions B s.e. Z-value p Perceived Ability Employee task performance 0.05 0.03 1.53 0.13 Employee voice -0.01 0.02 -0.52 0.60 Perceived Benevolence Employee task performance 0.05 0.03 1.46 0.14 Employee voice 0.03 0.02 1.18 0.24 Perceived Integrity Employee task performance 0.08* 0.03 2.93 0.00 Employee voice -0.03 0.02 -1.49 0.14 Abusive Supervision Perceived ability 0.03 0.03 1.07 0.28 Perceived benevolence -0.04 0.03 -1.39 0.16 Perceived integrity -0.04† 0.02 -1.76 0.08 R-Squared Perceived Ability 0.03 Perceived Benevolence 0.05 Perceived Integrity 0.04 Abusive Supervision 0.09 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 86 Table 32 Path model for trust as three factors mediating the relationship between employee task performance and voice behavior with initiating structure. Regressions B s.e. Z-value p Perceived Ability Employee task performance 0.06* 0.03 2.03 0.04 Employee voice -0.02 0.02 -0.92 0.36 Perceived Benevolence Employee task performance 0.05† 0.03 1.76 0.08 Employee voice 0.02 0.02 0.76 0.45 Perceived Integrity Employee task performance 0.08* 0.03 3.24 0.00 Employee voice -0.04* 0.02 -1.99 0.05 Initiating Structure Perceived ability 0.52* 0.23 2.29 0.02 Perceived benevolence 0.04 0.25 0.16 0.87 Perceived integrity 0.24 0.23 1.06 0.29 R-Squared Perceived Ability 0.02 Perceived Benevolence 0.03 Perceived Integrity 0.05 Initiating Structure 0.08 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 87 Table 33 Path model for trust as three factors mediating the relationship between employee task performance and voice behavior with consideration behavior. Regressions B s.e. Z-value p Perceived Ability Employee task performance 0.05† 0.03 1.73 0.08 Employee voice 0.00 0.02 -0.10 0.92 Perceived Benevolence Employee task performance 0.05 0.03 1.45 0.15 Employee voice 0.04† 0.02 1.65 0.10 Perceived Integrity Employee task performance 0.08* 0.03 3.11 0.00 Employee voice -0.03 0.02 -1.51 0.13 Consideration Behavior Perceived ability 0.12 0.21 0.58 0.56 Perceived benevolence 0.77* 0.17 4.46 0.00 Perceived integrity 0.02 0.20 0.09 0.93 R-Squared Perceived Ability 0.03 Perceived Benevolence 0.05 Perceived Integrity 0.04 Consideration Behavior 0.52 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 88 Table 34 Path model for trust as three factors mediating the relationship between employee task performance and voice behavior with ethical leadership. Regressions B s.e. Z-value p Perceived Ability Employee task performance 0.03 0.03 1.13 0.26 Employee voice 0.00 0.02 -0.08 0.94 Perceived Benevolence Employee task performance 0.04 0.03 1.19 0.23 Employee voice 0.03 0.02 1.47 0.14 Perceived Integrity Employee task performance 0.07* 0.03 2.39 0.02 Employee voice -0.03 0.02 -1.25 0.21 Ethical Leadership Perceived ability 0.30 0.26 1.14 0.26 Perceived benevolence 0.56* 0.22 2.52 0.01 Perceived integrity 0.07 0.24 0.31 0.75 R-Squared Perceived Ability 0.01 Perceived Benevolence 0.03 Perceived Integrity 0.03 Consideration Behavior 0.08 Note: * indicates p < .05. 89 Table 35 Path model for trust as three factors mediating the relationship between employee task performance and voice behavior with laissez-faire leadership. Regressions B s.e. Z-value p Perceived Ability Employee task performance 0.04 0.03 1.48 0.14 Employee voice 0.00 0.02 -0.07 0.94 Perceived Benevolence Employee task performance 0.04 0.03 1.51 0.13 Employee voice 0.03 0.02 1.33 0.18 Perceived Integrity Employee task performance 0.07* 0.03 2.79 0.01 Employee voice -0.02 0.02 -1.20 0.23 Laissez-faire Leadership Perceived ability -0.01 0.13 -0.07 0.94 Perceived benevolence -0.05 0.10 -0.51 0.61 Perceived integrity -0.13 0.10 -1.32 0.19 R-Squared Perceived Ability 0.03 Perceived Benevolence 0.05 Perceived Integrity 0.04 Laissez-faire Leadership 0.03 Note: * indicates p < .05. 90 Table 36 Path model for trust as three factors including the direct paths from predictors to leader behaviors. Regressions B s.e. Z-value p Perceived Ability Employee task performance 0.05 0.03 1.4 0.16 Employee voice -0.02 0.02 -0.74 0.46 Perceived Benevolence Employee task performance 0.04 0.04 1.15 0.25 Employee voice 0.04† 0.02 1.7 0.09 Perceived Integrity Employee task performance 0.07* 0.03 2.58 0.01 Employee voice -0.04† 0.02 -1.73 0.08 Consideration Behavior Perceived ability 0.16 0.16 1.01 0.31 Perceived benevolence 0.44* 0.17 2.63 0.01 Perceived integrity 0.06 0.18 0.32 0.75 Employee task performance 0.11* 0.06 1.95 0.05 Employee voice 0.37* 0.04 9.49 0.00 Initiating Structure Perceived ability 0.48* 0.22 2.17 0.03 Perceived benevolence -0.07 0.25 -0.30 0.77 Perceived integrity 0.31 0.23 1.38 0.17 Employee task performance 0.07 0.08 0.86 0.39 Employee voice 0.28* 0.07 3.92 0.00 Abusive Supervision Perceived ability 0.04 0.03 1.14 0.26 Perceived benevolence -0.05 0.03 -1.58 0.12 Perceived integrity -0.03 0.02 -1.39 0.16 Employee task performance -0.03 0.02 -1.59 0.11 Employee voice 0.02* 0.01 2.13 0.03 Laissez-faire Leadership Perceived ability 0.08 0.15 0.55 0.58 Perceived benevolence -0.19† 0.12 -1.68 0.09 Perceived integrity -0.02 0.10 -0.19 0.85 Employee task performance -0.01 0.06 -0.17 0.86 Employee voice 0.06 0.04 1.47 0.14 Ethical Leadership Perceived ability 0.24 0.21 1.13 0.26 Perceived benevolence 0.42* 0.21 2.03 0.04 Perceived integrity 0.16 0.27 0.58 0.56 Employee task performance 0.15† 0.08 1.88 0.06 Employee voice 0.34* 0.06 5.31 0.00 91 Table 36 (cont’d) R-Squared Perceived Ability 0.01 Perceived Benevolence 0.04 Perceived Integrity 0.04 Consideration Behavior 0.45 Initiating Structure 0.23 Abusive Supervision 0.11 Laissez-faire Leadership 0.04 Ethical Leadership 0.34 Note: * indicates p < .05; † indicates that the result is approaching significance (p < .10). 92 Table 37 Path model for trust as three factors with first- and second-stage moderation of the indirect pathway and moderation of the direct pathway. Regressions B s.e. Z-value p Perceived Ability Employee task performance 0.05 0.04 1.31 0.19 Employee voice 0.00 0.02 0.13 0.90 Positive reciprocity orientation 0.08 0.10 0.82 0.41 Task * pos. reciprocity ori. 0.03 0.06 0.47 0.64 Voice * pos. reciprocity ori. -0.11* 0.05 -2.30 0.02 Perceived Benevolence Employee task performance 0.03 0.04 0.88 0.38 Employee voice 0.05* 0.03 2.04 0.04 Positive reciprocity orientation 0.09 0.11 0.88 0.38 Task * pos. reciprocity ori. -0.06 0.07 -0.83 0.40 Voice * pos. reciprocity ori. -0.05 0.05 -0.87 0.39 Perceived Integrity Employee task performance 0.05† 0.03 1.67 0.09 Employee voice -0.02 0.02 -0.70 0.49 Positive reciprocity orientation 0.19* 0.09 2.11 0.04 Task * pos. reciprocity ori. 0.04 0.05 0.80 0.42 Voice * pos. reciprocity ori. -0.08† 0.04 -1.81 0.07 Consideration Behavior Perceived ability 0.15 0.19 0.81 0.42 Need for affiliation 0.14 0.21 0.67 0.50 Perceived ability * Need for affiliation -0.03 0.34 -0.09 0.93 Perceived benevolence 0.42* 0.18 2.40 0.02 Perceived benevolence * Need for affiliation 0.09 0.32 0.29 0.77 Perceived integrity 0.04 0.19 0.24 0.81 Perceived integrity * Need for affiliation -0.09 0.12 -0.72 0.47 Employee task performance 0.09 0.06 1.43 0.15 Employee voice 0.37* 0.04 9.33 0.00 Task * pos. reciprocity ori. 0.10 0.10 1.06 0.29 Voice * pos. reciprocity ori. -0.04 0.07 -0.48 0.63 93 Table 37 (cont’d) Regressions B s.e. Z-value p Initiating Structure Perceived ability 0.45* 0.23 1.93 0.05 Need for achievement 0.02 0.25 0.07 0.94 Perceived ability * Need for achievement 0.37 0.42 0.88 0.38 Perceived benevolence -0.04 0.26 -0.17 0.87 Perceived benevolence * Need for achievement -0.22 0.38 -0.57 0.57 Perceived integrity 0.28 0.23 1.22 0.22 Perceived integrity * Need for achievement 0.37 0.34 1.09 0.27 Employee task performance 0.06 0.08 0.79 0.43 Employee voice 0.31* 0.07 4.73 0.00 Task * pos. reciprocity ori. 0.24* 0.11 2.26 0.02 Voice * pos. reciprocity ori. -0.22* 0.11 -2.02 0.04 Abusive Supervision Perceived ability 0.01 0.03 0.29 0.77 Need for power 0.02 0.02 1.25 0.21 Perceived ability * Need for power 0.04 0.04 0.96 0.34 Perceived benevolence -0.03 0.03 -1.15 0.25 Perceived benevolence * Need for power -0.06* 0.03 -1.94 0.05 Perceived integrity -0.02 0.02 -0.90 0.37 Perceived integrity * Need for power -0.05 0.03 -1.54 0.12 Employee task performance -0.01 0.01 -1.03 0.30 Employee voice 0.01† 0.01 1.79 0.07 Task * pos. reciprocity ori. -0.01 0.01 -0.55 0.58 Voice * pos. reciprocity ori. 0.01 0.01 0.50 0.62 Laissez-faire leadership Perceived ability 0.00 0.15 0.03 0.98 Moral identity 0.03 0.09 0.33 0.74 Perceived ability * Moral identity 0.00 0.18 0.01 0.99 Perceived benevolence -0.10 0.10 -0.94 0.35 Perceived benevolence * Moral identity -0.09 0.12 -0.72 0.47 Perceived integrity 0.00 0.11 0.02 0.99 Perceived integrity * Moral identity 0.07 0.14 0.47 0.64 Employee task performance 0.04 0.04 0.99 0.32 Employee voice 0.03 0.03 1.03 0.31 94 Table 37 (cont’d) Task * pos. reciprocity ori. 0.04 0.05 0.78 0.43 Voice * pos. reciprocity ori. -0.01 0.05 -0.14 0.89 Ethical Leadership Perceived ability 0.31 0.25 1.25 0.21 Leadership identity 0.26 0.23 1.15 0.25 Perceived ability * Leadership identity -0.11 0.44 -0.25 0.80 Perceived benevolence 0.35 0.25 1.38 0.17 Perceived benevolence * Leadership identity 0.40 0.32 1.26 0.21 Perceived integrity 0.10 0.28 0.36 0.72 Perceived integrity * Leadership identity -0.08 0.32 -0.24 0.81 Employee task performance 0.13 0.09 1.50 0.13 Employee voice 0.32* 0.07 4.62 0.00 Task * pos. reciprocity ori. -0.07 0.12 -0.63 0.53 Voice * pos. reciprocity ori. 0.05 0.11 0.45 0.65 R-Squared Perceived Ability 0.01 Perceived Benevolence 0.04 Perceived Integrity 0.04 Consideration Behavior 0.45 Initiating Structure 0.23 Abusive Supervision 0.11 Laissez-faire Leadership 0.04 Ethical Leadership 0.34 95 Table 38 Indices of moderated mediation for the path model in table 37. Z- Variables B s.e. value p ci.lower ci.upper Moderator Moderator Predictor (1) Mediator (2) Outcome Pos. Recip. Need for Voice Orientation Ability Affiliation Consideration 0.00 0.04 0.08 0.93 -0.08 0.10 Pos. Recip. Need for Task Orientation Ability Affiliation Consideration 0.00 0.02 -0.04 0.97 -0.04 0.05 Pos. Recip. Need for Initiating - Voice Orientation Ability Achievement Structure 0.04 0.06 -0.71 0.48 -0.17 0.04 Pos. Recip. Need for Initiating Task Orientation Ability Achievement Structure 0.01 0.04 0.26 0.80 -0.04 0.11 Pos. Recip. Need for Abusive Voice Orientation Ability Power Supervision 0.00 0.01 -0.79 0.43 -0.02 0.00 Pos. Recip. Need for Abusive Task Orientation Ability Power Supervision 0.00 0.00 0.29 0.77 0.00 0.01 Pos. Recip. Moral Laissez-faire Voice Orientation Ability Identity Leadership 0.01 0.05 0.23 0.82 -0.08 0.13 Pos. Recip. Moral Laissez-faire Task Orientation Ability Identity Leadership 0.00 0.03 -0.10 0.92 -0.07 0.05 Pos. Recip. Leadership Ethical Voice Orientation Ability Identity Leadership 0.00 0.02 -0.19 0.85 -0.06 0.04 Pos. Recip. Leadership Ethical - Task Orientation Ability Identity Leadership 0.01 0.03 -0.18 0.86 -0.08 0.04 Pos. Recip. Need for Voice Orientation Benevolence Affiliation Consideration 0.01 0.03 0.35 0.73 -0.04 0.08 Pos. Recip. Need for Task Orientation Benevolence Affiliation Consideration 0.01 0.04 0.33 0.74 -0.05 0.11 Pos. Recip. Need for Initiating Voice Orientation Benevolence Achievement Structure 0.00 0.00 0.73 0.47 0.00 0.01 96 Table 38 (cont’d) Pos. Recip. Need for Initiating Task Orientation Benevolence Achievement Structure 0.00 0.01 0.66 0.51 -0.01 0.02 Pos. Recip. Need for Abusive - Voice Orientation Benevolence Power Supervision 0.02 0.03 -0.61 0.54 -0.09 0.04 Pos. Recip. Need for Abusive - Task Orientation Benevolence Power Supervision 0.02 0.04 -0.54 0.59 -0.13 0.04 Pos. Recip. Leadership Ethical Voice Orientation Benevolence Identity Leadership 0.01 0.01 0.62 0.53 -0.02 0.03 Pos. Recip. Leadership Ethical Task Orientation Benevolence Identity Leadership 0.00 0.01 -0.38 0.70 -0.02 0.01 Pos. Recip. Need for - Voice Orientation Integrity Affiliation Consideration 0.03 0.04 -0.83 0.41 -0.11 0.02 Pos. Recip. Need for Task Orientation Integrity Affiliation Consideration 0.01 0.03 0.49 0.63 -0.03 0.09 Pos. Recip. Need for Initiating Voice Orientation Integrity Achievement Structure 0.00 0.00 1.14 0.25 0.00 0.01 Pos. Recip. Need for Initiating Task Orientation Integrity Achievement Structure 0.00 0.00 -0.61 0.54 -0.01 0.00 Pos. Recip. Need for Abusive Voice Orientation Integrity Power Supervision 0.01 0.03 0.21 0.83 -0.05 0.08 Pos. Recip. Need for Abusive Task Orientation Integrity Power Supervision 0.00 0.02 -0.14 0.89 -0.05 0.04 Pos. Recip. Moral Laissez-faire Voice Orientation Integrity Identity Leadership 0.00 0.04 0.08 0.93 -0.08 0.10 Pos. Recip. Moral Laissez-faire Task Orientation Integrity Identity Leadership 0.00 0.02 -0.04 0.97 -0.04 0.05 Pos. Recip. Leadership Ethical - Voice Orientation Integrity Identity Leadership 0.04 0.06 -0.71 0.48 -0.17 0.04 Pos. Recip. Leadership Ethical Task Orientation Integrity Identity Leadership 0.01 0.04 0.26 0.80 -0.04 0.11 97 Table 39 Fit indices for the path model in table 37. χ² df CFI TLI RMSEA SRMR 326.25* 163.00 0.71 0.57 0.08 0.06 Note: * indicates p < .05. 98 Table 40 Simple slopes analysis of positive reciprocity orientation moderating employee voice with a leader’s perceived ability. See also: Figure 6. Slope S.E. t-value p-value SD +1 -0.04 0.02 -1.59 0.11 Mean 0.02 0.02 0.96 0.34 SD - 1 0.07* 0.03 2.74 0.01 Note: * indicates p < .05. Table 41 Simple slopes analysis of positive reciprocity orientation moderating employee voice with a leader’s perceived integrity. See also: Figure 7. Slope S.E. t-value p-value SD +1 -0.03 0.02 -1.18 0.24 Mean 0.00 0.02 0.05 0.96 SD - 1 0.03 0.03 1.14 0.26 Note: * indicates p < .05. Table 42 Simple slopes analysis for need for power (mod), leader’s perception of their perceived benevolence (x), and abusive supervision (y). See also: Figure 8. Slope S.E. t-value p-value SD +1 -0.11* 0.03 -3.58 < 0.01 Mean -0.05* 0.02 -2.16 0.03 SD - 1 0.01 0.03 0.52 0.61 Note: * indicates p < .05. 99 Table 43 Simple slopes analysis for need for power (mod), leader’s perception of their perceived benevolence (x), and abusive supervision (y). See also: Figure 9. Slope S.E. t-value p-value SD +1 -0.03 0.02 -1.18 0.24 Mean 0.00 0.02 0.05 0.10 SD - 1 0.03 0.03 1.14 0.26 Note: * indicates p < .05. Table 44 Summary of results from proposed hypotheses and exploratory analyses. Hypothesis Proposed Relationship Result H1 Task -> (+) Perceived Trust Supported H2 Voice -> (+) Perceived Trust Unsupported H1: Unsupported H2: Unsupported, but significant Positive Reciprocity Orientation in the opposite direction H3 moderates relationships in H1 and H2 (weakens relationship) Perceived Trust -> (+) Abusive H4 Supervision Unsupported Need for Power moderates the H5 relationship in H4 Unsupported Perceived Trust -> (-) Initiating Unsupported, but significant H6 Structure positive predictor Unsupported, but approaches significance in the opposite Need for Achievement moderates the direction (strengthens H7 relationship in H6 relationship) Perceived Trust -> (-) Consideration Unsupported, but significant H8 Behavior positive predictor Need for Affiliation moderates the H9 relationship in H8 Unsupported Unsupported, but significant H10a Perceived Trust -> (+) Incivility negative predictor Need for Power moderates the H10b relationship in H10a Unsupported 100 Table 44 (cont’d) Significant Exploratory Relationships from Path Models Relationship Significance p < .05; nonsignificant when moderator included (pos. rec. Voice -> (+) Abusive Supervision orientation). Voice -> (+) Initiating Structure p < .05 Voice -> (+) Consideration Behavior p < .05 Leadership Identity -> (+) Ethical p < .05; nonsignificant with trust Leadership as 3 factors p < .05; nonsignificant with trust as 3 factors or inclusion of Task -> (+) Ethical Leadership moderator Voice -> (+) Ethical Leadership p < .05 Perceived Trust -> (+) Ethical p < .05; nonsignificant with Leadership moderators at p < .10 p < .05; nonsignificant with Task -> (+) Ability more complicated models Task -> (+) Benevolence Nonsignificant at p < .10 Task -> (+) Integrity p < .05 Ability -> (+) Initiating Structure p < .05 Benevolence -> (+) Consideration p < .05 Benevolence -> (+) Ethical Leadership p < .05 Integrity -> (-) Abusive Supervision Nonsignificant at p < .10 Exploratory Moderation Results Moderator Relationship Effect on Relationship Pos. Recip. Ori. Task-> (+) Initiating Structure Strengthens Pos. Recip. Ori. Voice -> (+) Initiating Structure Weakens Pos. Recip. Weaker, nonsignificant at p < Ori. Voice -> (-) Perceived Trust .10 Pos. Recip. Ori. Voice -> (-)Perceived Ability Weakens Pos. Recip. Strengthens, nonsignificant at p Ori. Voice -> (-) Perceived Integrity < .10 Need for Power Benevolence -> (-) Abusive Supervision Weakens "->" indicates a relationship. (+) indicates the relationship is positive. (-) indicates the relationship is negative. 101 Figures Figure 1: Proposed Model 102 Figure 2: Moderation Effect of Positive Reciprocity Orientation on the Relationship between Employee Voice Behavior and a Leader’s Perceived Trust (See also: Table 8) 103 Figure 3: Moderation Effect of Need for Achievement on the Relationship between a Leader’s Perceived Trust and their Initiating Structure (See also: Table 11a) 104 Figure 4: Moderation Effect of Positive Reciprocity Orientation on the Relationship between an Employee’s Voice Behavior and a Leader’s Initiating Structure (See also: Table 18) 105 Figure 5: Moderation Effect of Positive Reciprocity Orientation on the Relationship between an Employee’s Task Performance and a Leader’s Initiating Structure (See also: Table 19) 106 Figure 6: Moderation Effect of Positive Reciprocity Orientation on the Relationship between an Employee’s Voice Behavior and a Leader’s Perceived Ability (See also: Table 40) 107 Figure 7: Moderation Effect of Positive Reciprocity Orientation on the Relationship between an Employee’s Voice Behavior and a Leader’s Perceived Integrity (See also: Table 41). 108 Figure 8: Moderation Effect of Need for Power on the Relationship between a Leader’s Perceived Benevolence and a Leader’s Abusive Supervision (See also: Table 42). 109 REFERENCES Ackermann, K. A., Fleiß, J., & Murphy, R. O. (2016). Reciprocity as an individual difference. Journal of Conflict Resolution, 60(2), 340-367. Anand, S., Hu, J., Liden, R. C., & Vidyarthi, P. R. (2011). Leader-member exchange: Recent research findings and prospects for the future. The Sage handbook of leadership, 311- 325. Aquino, K., & Reed, I. I. (2002). The self-importance of moral identity. Journal of personality and social psychology, 83(6), 1423. Barling, J., Slater, F., & Kelloway, E. K. (2000). Transformational leadership and emotional intelligence: An exploratory study. Leadership & Organization Development Journal. Bass, B. M., & Avolio, B. J. (2004). Multifactor Leadership Questionnaire: MLQ; manual and sampler set. Mind Garden. Bass, B. M., & Stogdill, R. M. (1990). Bass & Stogdill's handbook of leadership: Theory, research, and managerial applications. Simon and Schuster. Bass, B. M., & Avolio, B. J. (1997). Concepts of leadership. Leadership: Understanding the dynamics of power and influence in organizations, 3-22. Blanken, I., Van De Ven, N., & Zeelenberg, M. (2015). A meta-analytic review of moral licensing. Personality and Social Psychology Bulletin, 41(4), 540-558. Bernerth, J. B., Cole, M. S., Taylor, E. C., & Walker, H. J. (2018). Control variables in leadership research: A qualitative and quantitative review. Journal of Management, 44(1), 131-160. Billings, R. S., & Wroten, S. P. (1978). Use of path analysis in industrial/organizational psychology: Criticisms and suggestions. Journal of Applied Psychology, 63(6), 677–688. Botero, I. C., & Van Dyne, L. (2009). Employee voice behavior: Interactive effects of LMX and power distance in the United States and Colombia. Management Communication Quarterly, 23(1), 84-104. Brower, H. H., Lester, S. W., Korsgaard, M. A., & Dineen, B. R. (2009). A closer look at trust between managers and subordinates: Understanding the effects of both trusting and being trusted on subordinate outcomes. Journal of Management, 35(2), 327-347. Brown, S. L., & Eisenhardt, K. M. (1997). The art of continuous change: Linking complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative science quarterly, 1-34. Brown, M. E., Treviño, L. K., & Harrison, D. A. (2005). Ethical leadership: A social learning perspective for construct development and testing. Organizational behavior and human decision processes, 97(2), 117-134. Buhrmester, M., Kwang, T., & Gosling, S. D. (2016). Amazon's Mechanical Turk: A new source of inexpensive, yet high-quality data?. 110 Busch, H. (2018). Power motivation. Motivation and Action, (3). 335-368. Boyatzis, R. E. (2016). David C. McClelland. The Wiley Encyclopedia of Personality and Individual Differences, (4). Butler Jr, J. K., & Cantrell, R. S. (1984). A behavioral decision theory approach to modeling dyadic trust in superiors and subordinates. Psychological reports, 55(1), 19-28. Chen, X., Zhu, Z., & Liu, J. (2019). Does A Trusted Leader Always Behave Better? The Relationship Between Leader Feeling Trusted by Employees and Benevolent and Laissez-Faire Leadership Behaviors. Journal of Business Ethics, 1-20. Colquitt, J. A., Scott, B. A., & LePine, J. A. (2007). Trust, trustworthiness, and trust propensity: A meta-analytic test of their unique relationships with risk taking and job performance. Journal of applied psychology, 92(4), 909. Cortina, L. M., Magley, V. J., Williams, J. H., & Langhout, R. D. (2001). Incivility in the workplace: Incidence and impact. Journal of occupational health psychology, 6(1), 64. Cropanzano, R., Anthony, E. L., Daniels, S. R., & Hall, A. V. (2017). Social exchange theory: A critical review with theoretical remedies. Academy of Management Annals, 11(1), 479- 516. Cropanzano, R., & Mitchell, M. S. (2005). Social exchange theory: An interdisciplinary review. Journal of management, 31(6), 874-900. Depue, R. A., & Morrone-Strupinsky, J. V. (2005). A neurobehavioral model of affiliative bonding: Implications for conceptualizing a human trait of affiliation. Behavioral and Brain Sciences, 28(3), 313-349. Detert, J. R., Burris, E. R., Harrison, D. A., & Martin, S. R. (2013). Voice flows to and around leaders: Understanding when units are helped or hurt by employee voice. Administrative Science Quarterly, 58(4), 624-668. DeRue, D. S., & Ashford, S. J. (2010). Who will lead and who will follow? A social process of leadership identity construction in organizations. Academy of management review, 35(4), 627-647. DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2007). Between facets and domains: 10 aspects of the Big Five. Journal of personality and social psychology, 93(5), 880. Dirks, K. T., & Ferrin, D. L. (2002). Trust in leadership: Meta-analytic findings and implications for research and practice. Journal of applied psychology, 87(4), 611. Dunbar, N. E. (2004). Theory in progress: Dyadic power theory: Constructing a communication- based theory of relational power. Journal of Family Communication, 4(3-4), 235-248. Enders, C. K., & Bandalos, D. L. (2001). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural equation modeling, 8(3), 430-457. 111 Eisenberger, R., Lynch, P., Aselage, J., & Rohdieck, S. (2004). Who takes the most revenge? Individual differences in negative reciprocity norm endorsement. Personality and Social Psychology Bulletin, 30(6), 787-799. Forstmeier, W., & Schielzeth, H. (2011). Cryptic multiple hypotheses testing in linear models: overestimated effect sizes and the winner's curse. Behavioral ecology and sociobiology, 65(1), 47-55. Funder, D. C., & Colvin, C. R. (1991). Explorations in behavioral consistency: properties of persons, situations, and behaviors. Journal of personality and social psychology, 60(5), 773. Goldberg, L. R. (1995). ORI TECHNICAL REPORT Vol. 48 No. 1 March 2008. Health, 2. Gerstner, C. R., & Day, D. V. (1997). Meta-Analytic review of leader–member exchange theory: Correlates and construct issues. Journal of applied psychology, 82(6), 827. Höglinger, M., & Diekmann, A. (2017). Uncovering a Blind Spot in Sensitive Question Research: False Positives Undermine the Crosswise-Model RRT. Political Analysis, 25(1), 131-137. doi:10.1017/pan.2016.5 Hollander, E. P. (1958). Conformity, status, and idiosyncrasy credit. Psychological review, 65(2), 117. Hooper, D., Coughlan, J., & Mullen, M. R. (2008). Structural equation modelling: Guidelines for determining model fit. Electronic journal of business research methods, 6(1), 53-60. Hayes, A. F. (2015). An index and test of linear moderated mediation. Multivariate behavioral research, 50(1), 1-22. Heckhausen, J., & Heckhausen, H. (Eds.). (2008). Motivation and action (Vol. 22). New York: Cambridge University Press. Heckhausen, H. (2013). The anatomy of achievement motivation (Vol. 1). Academic Press. Iacobucci, D. (2010). Structural equations modeling: Fit indices, sample size, and advanced topics. Journal of consumer psychology, 20(1), 90-98. Johnson, R. E., Venus, M., Lanaj, K., Mao, C., & Chang, C. H. (2012). Leader identity as an antecedent of the frequency and consistency of transformational, consideration, and abusive leadership behaviors. Journal of Applied Psychology, 97(6), 1262. Judge, T. A., Piccolo, R. F., & Ilies, R. (2004). The forgotten ones? The validity of consideration and initiating structure in leadership research. Journal of applied psychology, 89(1), 36. Keith, M. G., Tay, L., & Harms, P. D. (2017). Systems perspective of Amazon Mechanical Turk for organizational research: Review and recommendations. Frontiers in Psychology, 8, 1359. Kenny, D. A. (1975). Cross-lagged panel correlation: A test for spuriousness. Psychological bulletin, 82(6), 887. 112 Kerr, S., & Schriesheim, C. (1974). Consideration, initiating structure, and organizational criteria—an update of Korman’s 1966 review. Personnel Psychology, 27(4), 555-568. Korman, A. K. (1966). “Consideration,”“Initiating Structure,” and Organizational Criteria‐a Review. Personnel Psychology, 19(4), 349-361. Lensvelt-Mulders, G. J., Hox, J. J., Van der Heijden, P. G., & Maas, C. J. (2005). Meta-analysis of randomized response research: Thirty-five years of validation. Sociological Methods & Research, 33(3), 319-348. Lin, S. H. J., Ma, J., & Johnson, R. E. (2016). When ethical leader behavior breaks bad: How ethical leader behavior can turn abusive via ego depletion and moral licensing. Journal of Applied Psychology, 101(6), 815. Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance in cross- sectional research designs. Journal of applied psychology, 86(1), 114. Mayer, R. C., & Davis, J. H. (1999). The effect of the performance appraisal system on trust for management: A field quasi-experiment. Journal of applied psychology, 84(1), 123. Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of management review, 20(3), 709-734. Mayer, R. C., & Gavin, M. B. (2005). Trust in management and performance: Who minds the shop while the employees watch the boss?. Academy of management journal, 48(5), 874- 888. Margerison, C. (1984). Chief executives' perceptions of managerial success factors.Journal of Management Development, 3(4), 47–60. https://doi.org/10.1108/eb051569 Markus, H., & Wurf, E. (1987). The dynamic self-concept: A social psychological perspective. Annual review of psychology, 38(1), 299-337. Martinko, M. J., Harvey, P., Brees, J. R., & Mackey, J. (2013). A review of abusive supervision research. Journal of Organizational Behavior, 34(S1), S120-S137. McClelland, D. C. (1965). Toward a theory of motive acquisition. American psychologist, 20(5), 321. McClelland, D. C., Koestner, R., & Weinberger, J. (1989). How do self-attributed and implicit motives differ?. Psychological review, 96(4), 690. Merritt, A. C., Effron, D. A., & Monin, B. (2010). Moral self‐licensing: When being good frees us to be bad. Social and personality psychology compass, 4(5), 344-357. Mitchell, M. S., & Ambrose, M. L. (2007). Abusive supervision and workplace deviance and the moderating effects of negative reciprocity beliefs. Journal of applied psychology, 92(4), 1159. Miller, D. T., & Effron, D. A. (2010). Psychological license: When it is needed and how it functions. In Advances in experimental social psychology (Vol. 43, pp. 115-155). Academic Press. 113 Miner, J. B. (2005). Organizational Behavior: Essential theories of motivation and leadership. one (Vol. 1). ME Sharpe. Morrison, E. W. (2014). Employee voice and silence. Annu. Rev. Organ. Psychol. Organ. Behav., 1(1), 173-197. Molm, L. D. (2003). Theoretical comparisons of forms of exchange. Sociological Theory, 21(1), 1-17. Nicholls, J. G. (1984). Achievement motivation: Conceptions of ability, subjective experience, task choice, and performance. Psychological review, 91(3), 328. Ormiston, M. E., & Wong, E. M. (2013). License to ill: The effects of corporate social responsibility and CEO moral identity on corporate social irresponsibility. Personnel Psychology, 66(4), 861-893. Parker, I. (Ed.). (1998). Social constructionism, discourse and realism. Sage. Paolacci, G., Chandler, J., & Ipeirotis, P. G. (2010). Running experiments on amazon mechanical turk. Judgment and Decision making, 5(5), 411-419. Podsakoff, P. M., MacKenzie, S. B., Moorman, R. H., & Fetter, R. (1990). Transformational leader behaviors and their effects on followers' trust in leader, satisfaction, and organizational citizenship behaviors. The leadership quarterly, 1(2), 107-142. Podsakoff, P. M., Todor, W. M., & Skov, R. (1982). Effects of leader contingent and noncontingent reward and punishment behaviors on subordinate performance and satisfaction. Academy of management journal, 25(4), 810-821. Preacher, K. J., & Coffman, D. L. (2006, May). Computing power and minimum sample size for RMSEA [Computer software]. Available from http://quantpsy.org/. Preacher, K. J., Curran, P. J., & Bauer, D. J. (2006). Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of educational and behavioral statistics, 31(4), 437-448. Pellegrini, E. K., & Scandura, T. A. (2008). Paternalistic leadership: A review and agenda for future research. Journal of management, 34(3), 566-593. Rousseau, D., Schalk, R., Schalk, M. J. D., & Schalk, R. (Eds.). (2000). Psychological contracts in employment: Cross-national perspectives. Sage. Rosseel Y (2012). “lavaan: An R Package for Structural Equation Modeling.” Journal of Statistical Software, 48(2), 1–36. http://www.jstatsoft.org/v48/i02/. Sachdeva, S., Iliev, R., & Medin, D. L. (2009). Sinning saints and saintly sinners: The paradox of moral self-regulation. Psychological science, 20(4), 523-528. Schoorman, D. F., Mayer, R. C., & Davis, J. H. (2016). Empowerment in veterinary clinics: The role of trust in delegation. Journal of Trust Research, 6(1), 76-90. 114 Shore, L. M., & Coyle‐Shapiro, J. A. M. (2003). New developments in the employee– organization relationship. Journal of Organizational Behavior: The International Journal of Industrial, Occupational and Organizational Psychology and Behavior, 24(5), 443-450. Simms, L. J., Goldberg, L. R., Roberts, J. E., Watson, D., Welte, J., & Rotterman, J. H. (2011). Computerized adaptive assessment of personality disorder: Introducing the CAT–PD project. Journal of personality assessment, 93(4), 380-389. Steffens, N. K., Haslam, S. A., Reicher, S. D., Platow, M. J., Fransen, K., Yang, J., ... & Boen, F. (2014). Leadership as social identity management: Introducing the Identity Leadership Inventory (ILI) to assess and validate a four-dimensional model. The leadership quarterly, 25(5), 1001-1024. Sverdlik, N., Oreg, S., & Berson, Y. (2022). The power of benevolence: The joint effects of contrasting leader values on follower‐focused leadership and its outcomes. Journal of Organizational Behavior.Tepper, B. J. (2000). Consequences of abusive supervision. Academy of management journal, 43(2), 178-190. Tepper, B. J., Henle, C. A., Lambert, L. S., Giacalone, R. A., & Duffy, M. K. (2008). Abusive supervision and subordinates' organization deviance. Journal of applied psychology, 93(4), 721. Tepper, B. J., Simon, L., & Park, H. M. (2017). Abusive supervision. Annual Review of Organizational Psychology and Organizational Behavior, 4, 123-152. Umphress, E. E., Bingham, J. B., & Mitchell, M. S. (2010). Unethical behavior in the name of the company: the moderating effect of organizational identification and positive reciprocity beliefs on unethical pro-organizational behavior. Journal of applied psychology, 95(4), 769. Van Eeden, R., Cilliers, F., & Van Deventer, V. (2008). Leadership styles and associated personality traits: Support for the conceptualisation of transactional and transformational leadership. South African Journal of Psychology, 38(2), 253-267. Vriend, T., Said, R., Janssen, O., & Jordan, J. (2020). The dark side of relational leadership: Positive and negative reciprocity as fundamental drivers of follower’s intended pro-leader and pro-self unethical behavior. Frontiers in Psychology, 11, 1473. Wåhlin-Jacobsen, C. D. (2020). Open or closed? A social interaction perspective on line managers’ reactions to employee voice. Management Communication Quarterly, 34(1), 32-57. Wei, F., & Si, S. (2013). Psychological contract breach, negative reciprocity, and abusive supervision: The mediated effect of organizational identification. Management and Organization Review, 9(3), 541-561. Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological review, 92(4), 548. Westland, J. C. (2010). Lower bounds on sample size in structural equation modeling. Electronic commerce research and applications, 9(6), 476-487. 115 Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and organizational commitment as predictors of organizational citizenship and in-role behaviors. Journal of management, 17(3), 601-617. Yoshikawa, K., Wu, C. H., & Lee, H. J. (2019). Generalized exchange orientation: Conceptualization and scale development. Journal of Applied Psychology. Zerbe, W. J., & Paulhus, D. L. (1987). Socially desirable responding in organizational behavior: A reconception. Academy of management review, 12(2), 250-264. Zhu, W., May, D. R., & Avolio, B. J. (2004). The impact of ethical leadership behavior on employee outcomes: The roles of psychological empowerment and authenticity. Journal of Leadership & Organizational Studies, 11(1), 16-26. 116 APPENDIX Employee’s Voice Behavior (LePine & Van Dyne, 1998) Please rate the number of times that [Employee] did the following things in the past week. 1. [Employee] developed and made recommendations to me concerning issues that affected their work. 2. [Employee] spoke up and encouraged others in the unit to get involved in issues that affected our work. 3. [Employee] communicated their opinions about work issues to myself and others in the work group, even if their opinions were different and they disagreed. 4. [Employee] kept well informed about issues at work where their opinion could be useful. 5. [Employee] spoke up to challenge me as their supervisor with ideas for new projects or changes in procedures at work. Scale Slider ranging from 0 to 10+. Non-whole number values are not options. 117 Employee Task Performance (William & Anderson, 1991) Please rate the number of times [Employee] did the following things in the last week: 1. Adequately completed assigned duties. 2. Fulfilled responsibilities specified in job description. 3. Performed tasks that are expected of them. 4. Met formal performance requirements of the job. 5. Engaged in activities that will directly affect their performance evaluations. 6. Neglected aspects of the job they are obligated to perform. * 7. Failed to perform essential duties.* Scale Slider ranging from 0 to 10+. Non-whole number values are not options. 118 Reciprocity orientation (Eisenberger et al., 2004) Negative 1. If someone dislikes you, you should dislike them. 2. If a person despises you, you should despise them. 3. If someone says something nasty to you, you should say something nasty back. 4. If a person wants to be your enemy, you should treat them like an enemy. 5. If someone treats me badly, I feel I should treat them even worse. 6. If someone has treated you poorly, you should not return the poor treatment. * 7. If someone important to you does something negative to you, you should do something even more negative to them. 8. A person who has contempt for you deserves your contempt. 9. If someone treats you like an enemy, they deserve your resentment. 10. When someone hurts you, you should find a way they won’t know about to get even. 11. You should not give help to those who treat you badly. 12. When someone treats me badly, I still act nicely to them. * 13. If someone distrusts you, you should distrust them. Positive 1. If someone does me a favor, I feel obligated to repay them in some way. 2. If someone does something for me, I feel required to do something for them. 3. If someone gives me a gift, I feel obligated to get them a gift. 4. I always repay someone who has done me a favor. 119 5. I feel uncomfortable when someone does me a favor that I know I won’t be able to return. 6. If someone sends me a card on my birthday, I feel required to do the same. 7. When someone does something for me, I often find myself thinking about what I have done for them. 8. If someone says something pleasant to you, you should say something pleasant back. 9. I usually do not forget if I owe someone a favor, or if someone owes me a favor. 10. If someone treats you well, you should treat the person well in return. Scale (original was 7 point) 1 – Strongly disagree 2 – Somewhat disagree 3 – Neither agree nor disagree 4 – Somewhat agree 5 – Strongly agree 120 Leader’s perceived trust from followers Trust (Mayer & Gavin, 2005) (α = .81) 1. If [Employee] had their way, they wouldn’t let me have any influence over issues that are important to them. * 2. [Employee] would be willing to let me have complete control over their future in this company. 3. [Employee] really wishes they had a good way to keep an eye on me. * 4. [Employee] would be comfortable giving me a task or problem which was critical to them, even if they could not monitor my actions. 5. If someone questioned my motives, [employee] would give me the benefit of the doubt. Ability, Benevolence, and Integrity (Mayer & Davis, 1999) Ability 1. [Employee] believes I am very capable of performing my job. 2. [Employee] believes I am successful at the things I try to do. 3. [Employee] believes I have much knowledge about the work that needs done. 4. [Employee] feels very confident about my skills. 5. [Employee] believes I have specialized capabilities that can increase our performance. 6. [Employee] believes I am well qualified. 121 Benevolence 1. [Employee] believes I am very concerned about their welfare. 2. [Employee] believes their needs and desires are very important to me. 3. [Employee] believes I would not knowingly do anything to hurt them. 4. [Employee] believes I really look out for what is important to them. 5. [Employee] believes I will go out of my way to help them. Integrity 1. [Employee] would say I have a strong sense of justice. 2. [Employee] never has to wonder whether I will stick to my word. 3. [Employee] believes I try hard to be fair in dealings with others. 4. [Employee] believes my actions and behaviors are not very consistent.* 5. [Employee] likes my values. 6. [Employee] thinks sound principles seem to guide my behavior. Scale 1 – Never 2 – Rarely 3 – Sometimes 4 – Very often 5 – Always 122 Abusive Supervision (Tepper, 2000; Tepper et al., 2008) In the last week I… 1. Ridiculed [Employee]. 2. Told [Employee] their thoughts or feelings are stupid. 3. Gave [Employee] the silent treatment. 4. Put [Employee] down in front of others. 5. Invaded [Employee] ‘s privacy. 6. Reminded [Employee] of their past mistakes and failures. 7. Didn’t give [Employee] credit for jobs requiring a lot of effort. 8. Blamed [Employee] to save myself embarrassment. 9. Broke promises I made to [Employee]. 10. Expressed anger at [Employee] when I am mad for another reason. 11. Made negative comments about [Employee] to others. 12. Was rude to [Employee]. 13. Did not allow [Employee] to interact with their coworkers. 14. Told [Employee] they’re incompetent. 15. Lied to [Employee]. Scale Slider ranging from 0 to 10+. Non-whole number values are not options. 123 Initiating Structure (Ohio State Leadership Studies, 1962) Please rate how often you did the following things in the last week (seven days). 1. I made my attitude clear to [Employee]. 2. I tried out new ideas with [Employee]. 3. I ruled with an iron hand. 4. I criticized [Employee] ‘s poor work. 5. I spoke to [Employee] in a manner not to be questioned. 6. I assigned [Employee] to complete tasks. 7. I scheduled [Employee] ‘s work to be done. 8. I maintained definite standards of performance for [Employee]. 9. I emphasized the meeting of deadlines to [Employee]. 10. I encouraged [Employee]‘s use of uniform procedures. 11. I made sure that [Employee]’s part in the organization was understood. 12. I asked [Employee] to follow standard rules and regulations. 13. I let [Employee] know what is expected of them. 14. I saw to it [Employee] was working up to capacity. 15. I saw to it the work of [Employee] was coordinated with group members. 124 Scale Slider ranging from 0 to 10+. Non-whole number values are not options. 125 Consideration Behavior (Ohio State Leadership Studies, 1962) 1. I did personal favors for [Employee]. 2. I did little things to make it pleasant for [Employee] to be a member of the group. 3. I was easy to understand when communicating with [Employee]. 4. I found time to listen to [Employee]. 5. I kept to myself. * 6. I looked out for the personal welfare of [Employee]. 7. I refused to explain my actions to [Employee]. * 8. I acted without consulting [Employee]. * 9. I backed up [Employee] in their actions. 10. I treated [Employee] as my equal. 11. I was willing to make changes for [Employee]. 12. I was friendly and approachable with [Employee]. 13. I made [Employee] feel at ease when talking with them. 14. I put suggestions made by [Employee] into operation. 15. I get [Employee]’s approval on important matters before going ahead. Scale Slider ranging from 0 to 10+. Non-whole number values are not options. 126 Need for power (Simms et al., 2011) Please rate the extent to which you agree with the following statements about yourself: 1. Boss people around. 2. Like having authority over others. 3. Insist that others do things my way. 4. Make demands on others. 5. Have a strong need for power. 6. Am known as a controlling person. Scale 1 – Strongly disagree 2 – Somewhat disagree 3 – Neither agree nor disagree 4 – Somewhat agree 5 – Strongly agree 127 Need for achievement (Goldberg, 1995) Please rate the extent to which you agree with the following statements about yourself: 1. Go straight for the goal. 2. Work hard. 3. Turn plans into actions. 4. Plunge into tasks with all my heart. 5. Do more than what’s expected of me. 6. Set high standards for myself and others. 7. Demand quality. 8. Am not highly motivated to succeed.* 9. Do just enough work to get by.* 10. Put little time and effort into my work.* Scale 1 – Strongly disagree 2 – Somewhat disagree 3 – Neither agree nor disagree 4 – Somewhat agree 5 – Strongly agree 128 Need for affiliation (DeYoung, Quilty, & Peterson, 2007) Please rate the extent to which you agree with the following statements about yourself: 1. Am interested in people. 2. Make people feel at ease. 3. Know how to comfort others. 4. Inquire about others’ well-being. 5. Take time out for others. 6. Make people feel welcome. 7. Show my gratitude. 8. Make others feel good. 9. Feel others’ emotions. 10. Am not really interested in others.* 11. Rarely smile.* Scale 1 – Strongly disagree 2 – Somewhat disagree 3 – Neither agree nor disagree 4 – Somewhat agree 5 – Strongly agree 129 Incivility (Cortina, Magley, Williams, & Langhout, 2001) During the past week at work, how many times did you do the following things to [Employee]? 1. Put [Employee] down or was condescending to [Employee]? 2. Paid little attention to [Employee]’s statement or showed little interest in their opinion? 3. Made demeaning or derogatory remarks about [Employee]? 4. Addressed [Employee] in unprofessional terms, either publicly or privately? 5. Ignored or excluded [Employee] from professional camaraderie? 6. Doubted [Employee]’s judgment on a matter over which they had responsibility? 7. Made unwanted attempts to draw [Employee]into a discussion of personal matters? Scale Slider ranging from 0 to 10+. Non-whole number values are not options. 130 Laissez-faire leadership (Bass & Avolio, 2004) Please rate how often you did the following things in the last week (seven days). 1. I avoided getting involved when important issues arose. 2. I was absent when needed. 3. I avoided making decisions. 4. I delayed responding to urgent questions. Scale Slider ranging from 0 to 10+. Non-whole number values are not options. 131 Ethical Leadership (Brown, Trevino, & Harrison, 2004) Items that are not behaviors are crossed out to prevent socially desirable responses. Please rate how often you did the following things in the last week (seven days). 1. I listened to what [Employee] had to say. 2. I disciplined [Employee] when they violated ethical standards. 3. I conducted my personal life in an ethical manner. 4. I had the interests of [Employee] in mind. 5. I made fair and balanced decisions. 6. I could be trusted. 7. I discussed business ethics or values with [Employee]. 8. I set an example of how to do things the right way in terms of ethics. 9. I defined success not just by results but also the way that they are obtained. 10. When making decisions, I asked “What is the right thing to do?” Scale Slider ranging from 0 to 10+. Non-whole number values are not options. 132 Moral identity (Aquino & Reed, 2002) Listed below are some characteristics that may describe a person: Caring, compassionate, fair, friendly, generous, hardworking, helpful, honest, kind, ruthless, selfish, distant. The person with these characteristics could be you or it could be someone else. For a moment, visualize in your mind the kind of person who has these characteristics. Imagine how that person would think, feel, and act. When you have a clear image of what this person would be like, answer the following questions. 1. It would make me feel good to be a person who has these characteristics. 2. Being someone who has these characteristics is an important part of who I am. 3. A big part of my emotional well-being is tied up in having these characteristics. 4. I would be ashamed to be a person who has these characteristics.* 5. Having these characteristics is not really important to me.* 6. Having these characteristics is an important part of my sense of self. 7. I strongly desire to have these characteristics. 8. I often wear clothes that identify me as having these characteristics. 9. The types of things I do in my spare time (e.g., hobbies) clearly identify me as having these characteristics. 10. The kinds of books and magazines that I read identify me as having these characteristics. 11. The fact that I have these characteristics is communicated to others by my membership in certain organizations. 133 12. I am actively involved in activities that communicate to others that I have these characteristics. Scale 1 – Strongly disagree 2 – Somewhat disagree 3 – Neither agree nor disagree 4 – Somewhat agree 5 – Strongly agree 134 Leader identity (Steffens, Haslam, Reicher, Platow, Fransen, Yang, Ryan, Jetten, Peters, & Boen, 2014) Identity prototypicality: ‘Being one of us’ (α = .96) 1. I embody what the group stands for. 2. I am representative of members of my group. 3. I am a model member of the group. * 4. I exemplify what it means to be a member of the team. Identity advancement: ‘Doing it for us’ (α = .89) 5. I promote the interests of members of the group. 6. I act as a champion for the group. * 7. I stand up for the group. 8. When I act, I have the group's interests at heart. Identity entrepreneurship: ‘Crafting a sense of us’ (α = .88) 9. I make people feel as if they are part of the same group. 10. I create a sense of cohesion within the group.* 11. I develop an understanding of what it means to be a member of the group. 12. I shape members' perceptions of the group's values and ideals. Identity impresarioship: ‘Making us matter’ (α = .92) 135 13. I devise activities that bring the group together. 14. I arrange events that help the group function effectively. 15. I create structures that are useful for group members.* Scale 1 – Not at all 2 – Slightly 3 – Moderately 4 – Very much 5 - Completely 136